GreatSmiles

Tooth Pain at Night: What to Do for Relief

Reviewed 2 September 2026. Fifty-two records, each pulled and re-checked against Europe PMC before publication: four Cochrane reviews (two on antibiotics for pulp and apical disease, two on analgesics for acute postoperative pain in children), the 2024 American Dental Association guideline on pharmacologic management of acute dental pain and its 2025 emergency-medicine summary, one international consensus statement on pulp therapies, eight randomised controlled trial reports, twelve further systematic reviews or meta-analyses (among them a network meta-analysis of 28 trials and 5 306 patients, and an umbrella review of eleven systematic reviews), ten cross-sectional and survey studies, three case reports, and pooled, mechanistic, claims-based, cohort and service-research records. Every number below is printed in the abstract of the record cited beside it. Written for the person sitting up in bed at 2 a.m. with a throbbing molar and for clinicians writing the “until you can be seen” part of an advice script. Not medical advice.

The short answers

Why a toothache is worse at 2 a.m.

Four mechanisms are documented in the retrieved records, and together they explain the shape of the problem — spontaneous pain, throbbing, worse lying flat, worse with hot and cold, not relieved by whatever is in the kitchen.

One: the pulp is a closed chamber. The hallmark of pulpitis pathogenesis is a sharp rise in intrapulpal pressure caused by vasodilation and oedema, and the experimental work cited above shows that this pressure is not a bystander: macrophages in inflamed pulp tissue become pressure-sensing cells, and the mechano-inflammatory signalling they start amplifies the inflammation further (Hu et al., Int Dent J 2026-06-01, PMID 42364415). Anything that increases head and pulpal blood volume — lying flat, bending, a hot bath, alcohol, exertion — pushes on a chamber that cannot expand. That is why the same molar that was bearable at 4 p.m. is unbearable at 2 a.m., and it is a mechanism, not a trial: no study randomised patients to an extra pillow, and this page does not pretend otherwise.

Two: the inflamed pulp is sensitised to temperature, not only to pressure. In irreversible pulpitis tissue, six of ten studied inflammatory genes were significantly up-regulated (CCL2 in 63.3% of samples, P = .001587; MMP9 60%; TNFα and IL6 56.7% each; CXCL8 and NOD2 53.3% each), and CXCL8 expression correlated with thermal hyper-response (r = 0.641) (Bhat et al., Int Dent J 2026-08-01, PMID 42217380). Practically: cold, which reduces blood flow and therefore pressure, tends to quiet an inflamed pulp for a while, whereas heat tends to worsen it. A patient who reports that a cold drink relieves the pain is describing a well-recognised pattern in irreversible pulpitis; that information belongs in the notes when you are seen.

Three: the pain that wakes you is the pain that triage systems rank first. In a cross-sectional study of 1 541 consecutive patients assessed at a dental hospital, eleven conditions produced dental pain, and the two most common were symptomatic irreversible pulpitis (427 patients, 27.7%) and symptomatic apical periodontitis (429, 27.8%). Those groups also scored highest: visual analogue scale 59.4 ± 24.4 and pain rating index 5.5 ± 2.8 for pulpitis, 56.1 ± 32.0 and 5.6 ± 2.9 for apical periodontitis, and pain between appointments scored 8.2 ± 0.5 on the rating index. Spontaneous pain, sleep disturbance, short bursts of pain, and painkillers not working were each significantly associated with higher pain scores (Wang et al., Int Dent J 2025-08-01, PMID 40609202). Read the other way: if you have night pain that pills do not touch, you are describing the highest-scoring group in that dataset, and you should say all four things when you call.

Four: the tooth is not the only thing that can hurt, and the differential matters at 2 a.m. Odontogenic pain — toothache — is the most common orofacial pain condition and accounts for the vast majority of pain presented to dental practitioners; but the clinical picture is occasionally unclear or contradictory, and in those situations the clinician should consider reasons other than the teeth, because the primary aim is to treat the underlying cause (Edwards et al., Br Dent J 2025-04-01, PMID 40217034). A headache, an ear that hurts only when swallowed, one-sided facial pain with a blocked nose after a flight, or a burning cheek with no swelling and no cold trigger, is a different pathway. If the pattern does not fit the four mechanisms above, the goal tonight is comfort and an examination, not escalating doses of ibuprofen.

What other people actually did before they called

Three independent datasets, on three continents, describe the same behaviour: dental pain is met with whatever is in the house, at the wrong dose, and with information obtained too late.

Before the dental emergency clinic: what 180 patients had already doneShare of 180 consecutive patients with dental pain, Copenhagen, 2023-2024 survey dataTook at least one analgesic82.8%Took paracetamol75.6%Took ibuprofen54.4%Knew the ibuprofen daily ceiling41.0%Knew the paracetamol daily ceiling39.0%Exceeded the recommended maximum9.0%085% of 180 patients
Source: Larsen SK, Markvart M, Søndenbroe R, Dalhoff K, Jensen SS, Basic & Clinical Pharmacology & Toxicology 2025;136(2):e14124, PMID 39825523 — a cross-sectional study of dental pain at two Copenhagen dental emergency clinics, where 180 consecutive patients answered questionnaires and interviews: 82.8% (n = 149) had taken at least one analgesic, paracetamol was used by 75.6% (n = 136) and ibuprofen by 54.4% (n = 98); 41% (n = 73) and 39% (n = 70) knew the recommended maximum daily dose of ibuprofen and paracetamol, and 9% (n = 15) had exceeded the maximum dosage. One source, one unit (percentage of the 180 patients). Opioids (10%) and other analgesics (11.1%) are omitted because they answer a different question. The transferable number is the last bar: the overdose risk in dental pain is created at the kitchen table, not in the pharmacy, and 54% of the medicines in this study had been obtained where guidance should have been available.

In Copenhagen, among 180 patients presenting to a dental emergency clinic, 82.8% had already taken an analgesic and 9% had exceeded the recommended maximum dosage; 54% of all analgesics used had been obtained at a place where guidance should have been available — pharmacies, dental clinics or hospitals — and only 39% (paracetamol) and 41% (ibuprofen) correctly identified the maximum daily dose (Larsen et al., Basic Clin Pharmacol Toxicol 2025-02-01, PMID 39825523). In a survey of 1 588 Saudi adults, 750 (47.2%) reported using analgesics for dental pain, ibuprofen being the most common (58%); herbal remedies were used by 10.8%, rinsing with salted water by 5.4%, and only 4.7% sought a medical or dental consultation, while 27.3% combined several relief methods; self-prescribing was associated with age (p = 0.005) and marital status (p = 0.017) (Alsnani et al., Cureus 2026-01-01, PMID 41743002). In France, self-medication prevalence was 51.2% among patients referred to an endodontic post-emergency unit after an emergency department visit for acute pulpal or periapical pain (43 patients) and 45.5% among patients referred for treatment under general anaesthesia because of dental anxiety (66 patients), with no socio-demographic or behavioural factor explaining the attitude (Le Texier et al., Int Dent J 2026-04-01, PMID 41643598).

Even the complaints themselves carry a number worth knowing. In a cross-sectional analysis of 659 posts about dental pain selected from 4 051 screened, fear was the most frequent emotion (29.7%) followed by sadness (19.1%); sentiment was most negative for pain complaints, advice-seeking, emergency access and post-treatment pain (p < 0.001); mentions of professional dental care (47.4%) outnumbered mentions of home remedies (1.1–17.0%) (Meisha et al., Saudi Dent J 2026-01-01, PMID 41571930). The fear is not laziness: the same literature that describes what patients swallow also describes why they have to.

The health-services numbers explain the delay. Among 342 French general practitioners surveyed, 68% saw at least one dental pain or dental infection consultation each week; over 90% frequently or very frequently heard patients explain that their dentist’s next available appointment was too far away, or that the dentist had advised them to seek medical care first; only one third felt confident managing those consultations, and 95% reported prescribing antibiotics often or systematically; 77% redirected patients to a dentist (Ambroise et al., Sante Publique 2026-01-01, PMID 42168090). Searches of the dental and wider literature for outcomes that should be used to evaluate care in adults with acute dental pain or infection returned 19 438 records, from which 27 studies spanning 1993 to 2020 were selected — and dentistry-specific studies reported only clinical outcomes, while patient-reported outcomes such as satisfaction appeared in the wider healthcare literature (Thompson et al., BMJ Open 2022-02-01, PMID 35190445). A 2026 analysis of two Brazilian national oral health surveys (37 519 participants in 2010, 40 720 in 2023) put self-reported dental pain in the previous six months at 23.3% (95% CI 21.3–25.4) and 20.4% (18.6–22.3), with inequalities by socioeconomic position largely unchanged over the thirteen years except among five-year-olds (Chisini et al., J Dent 2026-10-01, PMID 42320589).

The first six hours: what to take, when, and what it buys you

The following is what the retrieved evidence supports for a healthy adult with toothache and no contraindication to nonsteroidal anti-inflammatory drugs. It is written so that each line can be traced to a number, because the alternative — a list of folk doses — is how the Copenhagen patients ended up exceeding the maximum in 9% of cases (Larsen et al., Basic Clin Pharmacol Toxicol 2025-02-01, PMID 39825523).

Dose ceilings, and the arithmetic that keeps you out of trouble

Nothing on this page is a prescription, and the reason a page like this can safely say “the ceiling matters” is that the failure mode has been measured. In the Copenhagen dental emergency clinics, of all analgesics used, most were obtained at pharmacies, dental clinics or hospitals — places where guidance should be available — yet 9% (15 of 180) had exceeded the recommended maximum dosage, and awareness of the maximum daily dose was 39% for paracetamol and 41% for ibuprofen (Larsen et al., Basic Clin Pharmacol Toxicol 2025-02-01, PMID 39825523). In the Saudi survey, 47.2% self-prescribed, 27.3% used multiple relief methods at once, and 10.8% used herbal remedies on top (Alsnani et al., Cureus 2026-01-01, PMID 41743002). The pattern is the danger: the risk does not come from taking paracetamol, it comes from taking paracetamol plus “a combination tablet” plus “a cold and flu sachet”, each of which contains paracetamol.

Step What the retrieved records support Where the harm hides
1. Read every label for the same ingredient Paracetamol (acetaminophen) was the most used analgesic in dental pain patients — 75.6% of 180 — and only 39% knew its maximum daily dose (Larsen et al., Basic Clin Pharmacol Toxicol 2025-02-01, PMID 39825523) Combination cold, flu and “night” products duplicate the ingredient; the overdose is arithmetic, not recklessness
2. Add a nonsteroidal only if you can take one Nonsteroidal plus acetaminophen ranked first at 6–8 h after root canal treatment (MD −6.28, 95% CI −11.99 to −0.56) (Kelidari et al., BMC Oral Health 2026-01-01, PMID 41588508); in the ADA guideline, nonsteroidal drugs alone or with acetaminophen likely give better relief with a better safety profile than opioids (Carrasco-Labra et al., J Am Dent Assoc 2024-02-01, PMID 38325969) A 200 mg diclofenac patch matched the oral combination for pain but caused less gastric irritation (Khanal et al., Ann Med Surg (Lond) 2025-12-01, PMID 41377332) — the gastrointestinal tract is the cost centre, and stomach pain is not a reason to stop eating
3. Do not exceed, and do not borrow someone else’s prescription Nonsteroidal monotherapy carried a higher risk of adverse events than other non-opioids and than opioids alone in a network meta-analysis of 28 trials and 5 306 patients, at very low to low certainty, with a strong nocebo signal (Magesty et al., Med Oral Patol Oral Cir Bucal 2026-05-01, PMID 41578906) Borrowed opioids are the measurable harm: among US paediatric dental visits involving opioids, 10.1% had an overdose or persistent-use outcome (Liu et al., J Am Dent Assoc 2025-12-01, PMID 41081662)
4. Reassess at the six-hour mark Ketorolac’s advantage disappeared by 12–24 h (Ping et al., Clin Exp Dent Res 2026-04-01, PMID 41704089); NSAIDs premedication acted at 6 h while corticosteroids worked at 12–48 h (Mrini et al., J Clin Exp Dent 2025-12-01, PMID 41716365) If nothing has worked after two correctly timed doses, the problem is pressure inside a tooth, and no further pill fixes that
5. Book the definitive visit before you feel better Both pulpotomy and root canal treatment reduced pain through day 7 in the PROVE pooled analysis (410 randomised), with more relief and less analgesic use after root canal treatment (El Karim et al., Int Endod J 2026-06-01, PMID 42370524) The pain-free interval after a pulp starts to necrose is the most commonly missed window; patients return with swelling instead of pain (Arifin et al., Odontology 2025-04-01, PMID 39352670)

What you should not do tonight

Topicals, gels, clove oil and the things that are not nothing

The honest summary of non-pharmaceutical and plant-based options is that a few have real signals and none has a guideline behind it.

For plant-derived analgesics, a systematic review of 21 studies found the strongest signals for clove oil (eugenol), turmeric (curcumin), capsaicin, ginger powder, jidabokuippo, cannabis, propolis, green tea and pomegranate, with mechanisms ranging from inhibition of prostaglandin synthesis to modulation of inflammatory pathways and direct effects on nociceptors, applied in pulpitis, post-extraction pain, temporomandibular disorders and oral mucositis (Reddy et al., Int J Dent 2025-01-01, PMID 41322706). Eugenol is a genuine pharmacological agent — it is also the ingredient that burns oral mucosa when a soaked pellet is left in place overnight, and the review reports no guideline-grade recommendation for any of these agents.

For acupuncture, a systematic review and meta-analysis screened 1 672 publications and included 23, of which 11 randomised controlled trials (668 patients) reported on postoperative acute dental pain; acupuncture produced lower postoperative pain scores than sham (relative risk −0.77, 95% CI −1.52 to −0.03), and the authors state the limitation plainly: methodological shortcomings make further high-quality research necessary (Müller et al., Jpn Dent Sci Rev 2023-12-01, PMID 36950225).

For cannabis-derived products, the randomised evidence is thin and specific: in a placebo-controlled trial of 61 patients with moderate to severe toothache randomised to cannabidiol 10 mg/kg, 20 mg/kg or placebo and monitored for three hours, both cannabidiol groups had significant reductions in visual analogue scale pain relative to baseline and to placebo, with a maximum median reduction of 73% at 180 minutes, and the higher dose also improved bite force at 90 and 180 minutes without other psychoactive or mood effects (Chrepa et al., J Dent Res 2024-03-01, PMID 37910667). A single dose, three hours of observation, 61 patients — that is the entire clinical case for it, and it is not a reason to skip dental treatment.

Cold is the home measure with the most direct experimental support, but note where the cold was applied in the trials: inside the mouth, on the injection site or inside the tooth. In a 60-patient randomised trial of symptomatic irreversible pulpitis in lower first molars, cryotherapy added to the nerve block raised anaesthesia success from 65.3% to 79.3% and significantly reduced pain during access opening; in the other 60-patient trial, ten seconds of ice plus cold lignocaine in maxillary premolars improved onset, efficacy and injection pain. In the meta-analysis of intracanal cryotherapy, twelve trials showed a pain reduction confined to 6 and 12 hours (Katyal et al., Med Int (Lond) 2025-09-01, PMID 40747149) (Anjitha et al., J Conserv Dent Endod 2025-12-01, PMID 41438450) (Mohamed et al., Sci Rep 2026-04-01, PMID 42031860). One randomised trial of 90 patients also compared intracanal cryotherapy with dexamethasone irrigation after root canal preparation: at 6 hours the dexamethasone group was about 1.4 visual analogue scale units below control, at 12 hours dexamethasone and cryotherapy were similar and both below control, at 24 hours dexamethasone remained about 1.1 units below control, and at 48 and 72 hours there were no significant differences (Gürler et al., Quintessence Int 2026-06-01, PMID 41885126). Read as a set, these five records say: cold is an adjunct in the dental chair with an effect measured in hours, and it is worth an ice pack on the cheek at home without being anybody’s treatment.

Red flags: when tonight stops being a tooth problem

The same records that tell you pills can carry you to morning also define the point where they cannot. The emergency-medicine summary of the 2024 guideline was written for exactly those patients: people for whom definitive dental treatment is not immediately available, in emergency departments, hospitals and urgent care (Green et al., Am J Emerg Med 2025-03-01, PMID 39764906).

Pregnancy, breastfeeding and the “is any of this safe” question

This section exists because the fear of taking something is more common than the fear of leaving the infection untreated — and because both are measurable. In an online survey of 130 pregnant women (mean age 32), only 26.15% attended the dentist regularly during pregnancy; among those who experienced toothache, 51.52% used analgesics and 1.54% used antibiotics; 92.31% believed dental infections should be treated during pregnancy, yet 76.92% considered dental radiography unsafe and only 50% considered local anaesthesia safe; the number of pregnancies was associated with dental visits (p = 0.048), age with analgesic use (p = 0.018) and education with the perception of radiography safety (p = 0.013) (Atav et al., Healthcare (Basel) 2026-04-01, PMID 42121581).

Two things follow from those numbers, and they point in opposite directions. The 51.52% figure says that most pregnant patients with toothache self-treat — so a page like this one has an audience. The 1.54% figure says they almost never self-treat with antibiotics — which happens to be the right instinct for tonight’s pain, given the Cochrane findings quoted above (Agnihotry et al., Cochrane Database Syst Rev 2019-05-01, PMID 31145805) (Cope et al., Cochrane Database Syst Rev 2024-05-01, PMID 38712714). The 76.92% and 50% figures are the problem: avoiding examination and imaging because of a belief that they are unsafe in pregnancy converts a toothache into the cellulitis case above (Bassuel et al., Int Breastfeed J 2026-03-01, PMID 41814371). What this page can responsibly say is the same thing the guideline says about the drugs: the balance of benefit favours non-opioid analgesia, with nonsteroidal drugs alone or with acetaminophen first-line (Carrasco-Labra et al., J Am Dent Assoc 2024-02-01, PMID 38325969) (Green et al., Am J Emerg Med 2025-03-01, PMID 39764906) — and that in pregnancy the choice, the dose and the trimester are a clinician’s decision, made with an obstetric history in front of them, not a search result.

Children: the night waking that is not teething

Children’s dental pain is a sleep problem with a dental cause, and the relationship runs both ways. A narrative review of the paediatric literature concluded that dental caries, temporomandibular disorders and dentofacial deformities induce sleep disturbances in children and adolescents, while obstructive sleep apnoea, sleep bruxism and other sleep problems are strongly linked to oral health conditions, with altered oral microbial colonisation, immune impairment, persistent inflammation and chronic pain as the proposed pathways (Zhang et al., Clin Exp Dent Res 2025-08-01, PMID 40792606).

The behavioural detail that most often decides whether a six-month-old’s night pain becomes a two-year-old’s cavities is night feeding, and it has numbers: in a retrospective study of 150 children aged 12 to 72 months, co-sleeping was reported for 18.7%, the mean decayed-missing-filled-teeth score was 7.14 in co-sleepers versus 3.49 in non-co-sleepers, and in 89.3% of co-sleeping cases the child fell asleep with milk remaining in the mouth; co-sleeping was significantly associated with night-time breastfeeding on demand, higher carbohydrate consumption and inadequate oral hygiene (p < 0.05) (Bayón-Hernández et al., Medicine (Baltimore) 2026-05-01, PMID 42216372). That is a caries-risk finding, not a pain-relief finding — and it is the reason the answer to “my child screams when put down at night” is usually an examination, not a dose.

For the dose itself, use the weight-based paediatric evidence rather than the adult table. The ADA systematic review of paediatric acute dental pain found six trials with eight comparisons: the combination of acetaminophen and ibuprofen probably reduces pain intensity more than acetaminophen alone (mean difference −0.75 points, 95% CI −1.22 to −0.27, moderate certainty) and about the same as ibuprofen alone (−0.01, 95% CI −0.53 to 0.51, moderate certainty), while ibuprofen versus placebo was −0.19 (95% CI −0.58 to 0.21) and acetaminophen versus placebo −0.13 (95% CI −0.52 to 0.26), both low certainty, with very low certainty for adverse effects (Miroshnychenko et al., J Am Dent Assoc 2023-05-01, PMID 37105668). The Cochrane reviews give the context for those two drugs in children generally: for ibuprofen, 43 trials and 3 935 analysed children, with a probably-relevant reduction in child-reported pain intensity within two hours of the intervention (SMD −1.12, 95% CI −1.39 to −0.86, three studies, 259 children) and 37 of 43 studies judged at high or unclear risk of bias (Pessano et al., Cochrane Database Syst Rev 2024-01-01, PMID 38180091); for diclofenac, 32 trials in 2 250 children, all under general anaesthesia, with none of the placebo comparisons reporting child-reported pain relief or pain intensity at all, and nausea or vomiting at a risk ratio of 0.83 (95% CI 0.38 to 1.80) (Ringsten et al., Cochrane Database Syst Rev 2023-12-01, PMID 38078559). Two lessons: the evidence for children is thinner than the confidence of the advice, and a drug with a placebo-controlled dental trial behind it (ibuprofen) is a different proposition from one borrowed from surgery (diclofenac).

And the safety warning is specific, not generic: among US paediatric dental visits from 2014 through 2019, 0.72% involved a high-risk medication (benzodiazepine, barbiturate or opioid) and 4.3% of those children had hospitalisation, an emergency department or urgent-care visit within seven days; odds were higher at ages 9 to 11 (OR 1.56), in children with complex chronic conditions (OR 2.22) and in hospital or ambulatory-surgery settings (OR 2.20); among visits involving opioids, 10.1% had an opioid-attributable outcome within seven days to a year, with the highest odds at ages 4 to 5 (OR 1.48) (Liu et al., J Am Dent Assoc 2025-12-01, PMID 41081662). If a prescription opioid is on the counter from someone else’s surgery, the arithmetic of your child’s age group is the one that looks worst.

When you call: how triage sorts 2 a.m. pain, and how fast it should be

The records here are few but directly relevant, because they describe the system rather than the tooth. In a Brazilian study of a dental emergency service, a reception and risk-classification protocol built on the Manchester Triage System was developed and validated with nine items — personal data, history of the present condition, pain, pain classification, vital signs, need for medication, medical history, priority and risk classification — then piloted in 80 patients: 93.75% reported pain, predominantly spontaneous and intense, and the most frequent classifications were yellow risk (50%) and green risk (45%); the average time between arrival and treatment was 1 hour 33 minutes, with a 2.5% dropout rate, and agreement between the ten validating professionals was 97% (Kappa) (Oliveira de Sousa et al., Front Health Serv 2026-01-01, PMID 42388910). That is what a functioning dental triage looks like: almost everyone who arrives in pain gets sorted into two categories, and the interval to being seen is measured in tens of minutes, not days.

The temporal pattern is worth knowing before you decide which door to walk through. In a retrospective study of 510 adult dental emergency presentations in western Romania, endodontic and periapical pathology was the predominant diagnostic macrocategory, with most cases being acute apical periodontitis and pulpitis; most patients presented in the evening (46.47%) and on weekends — Sundays 21% (n = 107) and Saturdays 16% (n = 82); and patients who came in the morning and afternoon were significantly older than those presenting in the evening (p = 0.009) or at night (p = 0.047) (Negru et al., Healthcare (Basel) 2026-05-01, PMID 42194458). Two readings of that dataset, and both are useful: you are not being dramatic, since evening and weekend dental emergencies are the modal presentation; and the people who wait until the morning are older, which in a service-research context usually means a different threshold for seeking care rather than a milder disease.

What you should say when you call, in the order triage tools ask for it: spontaneous pain or pain only when lying down; how long each episode lasts; whether analgesics worked at all — because in the 1 541-patient study, “taking painkillers was invalid” was itself one of the features associated with higher pain scores, alongside sleep disturbance and short-duration pain (Wang et al., Int Dent J 2025-08-01, PMID 40609202); presence and location of swelling; temperature; whether you can swallow and open normally; and any blood thinner, chemotherapy, transplant drug or cardiac condition (Gazal et al., Saudi J Anaesth 2026-01-01, PMID 41710634). Triage is a language, and the sentence that gets you an earlier slot is the one that uses it.

What the dentist can do on the first visit

The reason every section above ends with “book it” is that the definitive treatments have effect sizes and durability that analgesics do not, and because the guidelines are unambiguous that they are the point.

For a tooth with symptomatic irreversible pulpitis, the choice is now genuinely between two vital treatments and a root canal, and the best evidence is a pooled analysis of randomised trials conducted in eight countries with a standardised protocol: 410 patients randomised, data available for 385, pain assessed at baseline, day 3 and day 7, with analgesic consumption and early failure (need for additional intervention within four weeks) as outcomes; pain intensity fell significantly over time in each treatment group, and root canal treatment gave significantly more pain relief and lower analgesic intake than pulpotomy, while early failure was low in both (El Karim et al., Int Endod J 2026-06-01, PMID 42370524). An umbrella review of pulpotomy in mature permanent teeth frames the optimism more carefully: from 332 articles, 11 systematic reviews were extracted, only one rated moderate quality on AMSTAR 2 with most rated low to very low; pooled success rates ranged from 37% to 100% depending on the capping material, with Biodentine highest, and short-term success of 80% to 95% exceeded root canal treatment while long-term outcomes remained comparable — with the review’s own condition attached, that success depends on pre- and perioperative conditions, achieving haemostasis, and a definitive coronal restoration (Mtalsi et al., Cureus 2026-03-01, PMID 41948253). A 24-month follow-up of a randomised trial of 62 patients aged 14 to 60 (32 control, 30 with adjunctive cryotherapy irrigated for five minutes with saline cooled to 2.5 °C, mineral trioxide aggregate cap, glass-ionomer liner and definitive adhesive restoration) reported clinical and radiographic outcomes at 1, 3, 6, 12 and 24 months, with success defined as absence of spontaneous pain, tenderness to percussion or palpation, radiographic periapical pathosis, or need for root canal treatment (Gürler et al., Clin Oral Investig 2026-08-01, PMID 42565862).

For a baby tooth, the same logic, with a shorter list: the international consensus statement from sixteen specialists convened at the third IAPD Summit in Porto in November 2024 produced 38 evidence-based recommendations across pulp inflammation and diagnosis, caries excavation, management of pulpitis in primary and permanent teeth, and traumatic injuries, using a 7-point Likert Delphi process in which recommendations needed more than 70% consensus, and key findings emphasised selective caries removal over complete excavation, calcium silicate cements as the preferred materials for vital pulp therapy, and conservative approaches for irreversible pulpitis, with decision trees for implementation (Dhar et al., Int J Paediatr Dent 2026-03-01, PMID 41622922).

And if the tooth cannot be saved, removal is definitive in the literal sense: it is the only option in the Cochrane pulpitis review described as relieving the pain by removing the source (Agnihotry et al., Cochrane Database Syst Rev 2019-05-01, PMID 31145805). The after-care decisions that follow — what to eat, when to stop smoking or vaping, how to avoid a dry socket — have their own pages on this site, with their own numbers.

How to read this like a clinician

Four habits of reading, each of which changes what you would write on an advice sheet.

Distinguish the outcome from the intervention. The strongest numbers on this page (the PROVE pooled analysis (El Karim et al., Int Endod J 2026-06-01, PMID 42370524), the premedication meta-analysis (Mrini et al., J Clin Exp Dent 2025-12-01, PMID 41716365), the cryotherapy trials (Katyal et al., Med Int (Lond) 2025-09-01, PMID 40747149) (Anjitha et al., J Conserv Dent Endod 2025-12-01, PMID 41438450) (Gürler et al., Quintessence Int 2026-06-01, PMID 41885126) (Mohamed et al., Sci Rep 2026-04-01, PMID 42031860)) are all measured in patients who have had a dental procedure. Applying them to a person at home at 2 a.m. is a transfer, not a prescription; the closest thing to a natural experiment in the community data is the observation that painkillers being ineffective is itself a marker of pulpal severity (Wang et al., Int Dent J 2025-08-01, PMID 40609202).

Read the anaesthesia literature as a pain-management literature. The reason an emergency pulpitis appointment is uncomfortable is that the inferior alveolar nerve block frequently fails to provide profound pulpal anaesthesia; a narrative review of clinical trials and systematic reviews from 1975 to 2025 concludes that increasing anaesthetic volume and applying supplemental intraligamentary, intraosseous or intrapulpal injections significantly enhances anaesthesia success, with cryotherapy and warming or cooling of the solution as further adjuncts, and that intrapulpal anaesthesia remains essential when other techniques fail (Eldafrawi et al., Pain Res Manag 2026-01-01, PMID 41808844). That is a set of options your clinician has and you do not — another argument against waiting for a “better pill”.

Watch the certainty column, not the direction column. Nearly every comparison on this page has a GRADE rating attached, and the pattern is instructive: moderate certainty for combining acetaminophen with ibuprofen in children (Miroshnychenko et al., J Am Dent Assoc 2023-05-01, PMID 37105668) and for bupivacaine reducing analgesic consumption versus lidocaine with epinephrine (mean difference −1.91 doses, 95% CI −3.35 to −0.46) and versus mepivacaine (−1.58, 95% CI −2.21 to −0.95) (Miroshnychenko et al., J Am Dent Assoc 2023-01-01, PMID 36608963); low certainty for benzocaine topicals in acute irreversible pulpitis (Miroshnychenko et al., J Am Dent Assoc 2023-01-01, PMID 36608963), for antibiotics in pulpitis (Agnihotry et al., Cochrane Database Syst Rev 2019-05-01, PMID 31145805) and for ibuprofen versus placebo in children (Miroshnychenko et al., J Am Dent Assoc 2023-05-01, PMID 37105668); very low to low, with significant inconsistency, for the adverse-event rankings in the network meta-analysis of single-dose analgesics (Magesty et al., Med Oral Patol Oral Cir Bucal 2026-05-01, PMID 41578906); very low for the pooled prevalence of inappropriate antibiotic prescribing for pulpitis, where I² was 98% (Delgado-Giugni et al., Antibiotics (Basel) 2025-12-01, PMID 41594051). Local guidelines add one more variable: adapting a guideline to a country’s available and affordable medicines is itself a research project — the Iranian exercise that screened 763 articles and guidelines, retained 36 sources, and produced two flowcharts for mild, moderate and severe pain reached content validity indices of 0.99 at scale level with a lowest item-level index of 0.93 (Farahmand et al., BMC Oral Health 2026-03-01, PMID 41832479).

Expect the practice to be the weak link. Fewer than 3% of surveyed patients reported adverse effects from pain medication, and 81% were managed on non-opioids — the pharmacology is not what goes wrong (Polk et al., BMC Oral Health 2026-03-01, PMID 41845294). What goes wrong is access and habit: 95% of surveyed general practitioners prescribing antibiotics often or systematically for dental complaints, while only a third felt able to manage them (Ambroise et al., Sante Publique 2026-01-01, PMID 42168090); 19.2% of dentists prescribing antibiotics for a disease that needs a drill (Delgado-Giugni et al., Antibiotics (Basel) 2025-12-01, PMID 41594051); patients exceeding maximum analgesic doses while buying the drugs at a pharmacy that could have told them not to (Larsen et al., Basic Clin Pharmacol Toxicol 2025-02-01, PMID 39825523). Every recommendation on this page is cheap. The one thing that reliably fixes it is an appointment.

Evidence at a glance

Question Estimate as printed Design and caveat
Why is it worse lying down? Pulpal pressure correlates with severity; macrophage pressure-sensing r = 0.672 ± 0.105 at 24 h and 0.805 ± 0.077 at 72 h; 180 kPa in vitro amplified TRAF6-NF-κB and YAP signalling Animal and in-vitro model, n = 6 per group per time point; mechanism, not a human trial (Hu et al., Int Dent J 2026-06-01, PMID 42364415)
Is the pain score real? SIP VAS 59.4 ± 24.4 and SAP 56.1 ± 32.0 among 1 541 patients; sleep disturbance, spontaneous pain, short episodes and ineffective analgesics all associated with higher scores Cross-sectional hospital series, 11 conditions, all p < 0.05 for those associations (Wang et al., Int Dent J 2025-08-01, PMID 40609202)
Does ibuprofen plus acetaminophen work better? MD −6.28 (95% CI −11.99 to −0.56) on VAS at 6–8 h for the strongest combination after root canal treatment; oral paracetamol 500 + ibuprofen 400 matched a diclofenac patch for pain (3.06 vs 3.46 at 12 h) with more gastric irritation Network meta-analysis of 5 pooled trials (347 patients) plus one randomised trial of 108 patients; extraction and endodontic populations (Kelidari et al., BMC Oral Health 2026-01-01, PMID 41588508) (Khanal et al., Ann Med Surg (Lond) 2025-12-01, PMID 41377332)
Do opioids help more? No: the guideline panel found a beneficial net balance favouring non-opioids; naproxen with codeine gave an 83.44% VAS reduction versus 91.52% for dexketoprofen with paracetamol 2024 guideline with four systematic reviews (Carrasco-Labra et al., J Am Dent Assoc 2024-02-01, PMID 38325969); randomised trial of 62 patients with pericoronitis (Doğru et al., Sci Rep 2025-10-01, PMID 41107344)
Are the side effects of non-steroidals overrated? Non-steroidal monotherapy ranked worse for adverse events than other non-opioids and opioids alone, but few comparisons beat placebo and placebo event rates were high — read as nocebo; certainty very low to low Network meta-analysis of 28 RCTs, 5 306 patients, third-molar surgery (Magesty et al., Med Oral Patol Oral Cir Bucal 2026-05-01, PMID 41578906)
Will a topical gel get me through the night? Benzocaine 10% and 20% may increase the number of people with pain reduction versus placebo in acute irreversible pulpitis, described over 20–30 minutes Five trials, low certainty (Miroshnychenko et al., J Am Dent Assoc 2023-01-01, PMID 36608963); prolonged over-the-counter benzocaine use has produced methaemoglobinaemia requiring methylene blue (Panossian et al., Cureus 2026-03-01, PMID 41970091)
Does clove oil do anything? Among 21 included studies, eugenol, curcumin, capsaicin, ginger, propolis, green tea and pomegranate showed analgesic signals, via prostaglandin inhibition and nociceptor modulation Systematic review, GRADE-assessed, no guideline-grade recommendation for any agent (Reddy et al., Int J Dent 2025-01-01, PMID 41322706)
Is acupuncture or cannabidiol worth trying instead? Acupuncture versus sham: RR −0.77 (95% CI −1.52 to −0.03) in 11 trials (668 patients). CBD versus placebo: maximum median VAS reduction 73% at 180 min in 61 patients Small, heterogeneous, single-dose; the authors of both call for higher-quality trials (Müller et al., Jpn Dent Sci Rev 2023-12-01, PMID 36950225) (Chrepa et al., J Dent Res 2024-03-01, PMID 37910667)
Can an antibiotic be prescribed tonight? It can, and it will not help the pain: identical median pain scores in the one pulpitis trial (40 patients), no difference in three apical-infection trials (134 patients), and no trial of antibiotics without a procedure at all Two Cochrane reviews; low to very low certainty (Agnihotry et al., Cochrane Database Syst Rev 2019-05-01, PMID 31145805) (Cope et al., Cochrane Database Syst Rev 2024-05-01, PMID 38712714)
How common is wrong-dose self-treatment? 9% of 180 dental emergency patients exceeded the recommended maximum; 39% and 41% knew the daily ceiling for paracetamol and ibuprofen Questionnaires and interviews in two clinics, Copenhagen (Larsen et al., Basic Clin Pharmacol Toxicol 2025-02-01, PMID 39825523)
Does the child’s sleep improve after treatment? Quality-of-life scores improved after endodontic treatment in both pulpitis and necrosis groups (p < 0.001) with no between-group difference; patients’ expectations of treatment were unchanged afterwards Prospective cohort, 131 patients, one dental hospital in Japan (Arifin et al., Odontology 2025-04-01, PMID 39352670)

What the evidence does not support

Cost, coverage and the Utah part

Utah’s published children’s dental benefit is written around exactly the events this page describes. Under “Care of abscesses”, emergency treatment for an abscess or acute infection does not require prior authorisation, and the document says this includes services such as an emergency exam, diagnostic x-rays and the incision and drainage of the abscess to relieve pain and infection; “Emergency room services provided by a dentist” is listed as covered. Root canals on permanent teeth do not require prior authorisation and are covered when medically necessary to save a tooth, with third molars generally not covered; pulpotomy — the document’s “mini” root canal for baby teeth — is covered for infected baby teeth once per tooth without prior authorisation, but not for teeth already loose or too damaged to be restored. Simple extractions (routine removal, local anaesthesia, any necessary stitches) and surgical extractions (where bone must be removed or the tooth sectioned) likewise need no prior authorisation when the tooth cannot be saved by a filling or root canal. Two bitewing x-rays are allowed per calendar year, and when a full set is billed, no additional individual x-rays are covered for that visit (data as of 4 February 2026; see Additional documents).

What the plan will not do is pay for the things people imagine will make the wait bearable: analgesia with nitrous oxide is listed as a non-covered service — the document explains that while a dentist may use it to help a patient relax, it is not a separately reimbursable benefit and the cost is not covered by Medicaid; Inpatient Hospital Services are “No” under the dental plan, with hospital-related costs for dental procedures performed as an inpatient billed to the medical plan; general anaesthesia and intravenous conscious sedation are covered only where a child cannot be treated safely under local anaesthesia because of a physical or mental disability or another complex medical condition, documented in the record; behaviour management is not covered, and orally administered sedation medicines run through the Medicaid pharmacy benefit by prescription only. There is no line in the dental plan for a telephone advice call, and nothing in it about over-the-counter analgesics — which is why the practical Utah answer to a 2 a.m. toothache is the emergency exam, the x-rays and the drainage that are covered without authorisation, plus the pharmacy arithmetic in the section above. On price: an extract of modelled CMS allowed amounts for sixteen Wasatch Front ZIP codes, cached in this project’s research files on 31 August 2026, returns the same figures for every ZIP code in the set (about $89 at the modelled midpoint, with $58 and $174 as the low and high bounds), so the distance you live from a clinic is not what moves your bill — what moves it is how many surfaces a filling needs, whether a crown or a pulp treatment is added, and whether sedation is involved. Ask for the procedure code and the estimate in writing; the estimate you get from a Utah clinic for a coded emergency exam is comparable to the modelled midpoint above, not to its upper bound.

A plan for tonight and tomorrow morning

  1. Now. Sit upright for twenty minutes, cool (not ice against tissue) drink on the painful side if cold relieves it, one correctly dosed analgesic from what is safe for you, and nothing that duplicates an ingredient already taken today (Larsen et al., Basic Clin Pharmacol Toxicol 2025-02-01, PMID 39825523) (Hu et al., Int Dent J 2026-06-01, PMID 42364415).
  2. In the same hour. Write down: when it started, what makes it worse, whether lying down matters, whether the last dose helped at all — the four features triage asks about and the four the pain-score study found discriminative (Wang et al., Int Dent J 2025-08-01, PMID 40609202) (Oliveira de Sousa et al., Front Health Serv 2026-01-01, PMID 42388910).
  3. Before you go back to sleep. Check the red-flag list. Anything on it, go now (Green et al., Am J Emerg Med 2025-03-01, PMID 39764906) (Cope et al., Cochrane Database Syst Rev 2024-05-01, PMID 38712714).
  4. First thing. Call and say the words “spontaneous pain, wakes me at night, analgesics not helping”. Then ask for whatever the earliest slot is, and say what the alternative is — in the French survey, 95% of general practitioners would prescribe antibiotics for the same complaint, which tells you what a no-appointment answer tends to become (Ambroise et al., Sante Publique 2026-01-01, PMID 42168090).
  5. Until you are seen. Keep dosing on the schedule you set, not on the pain’s schedule; eat on the other side; do not apply heat; do not start leftovers of antibiotics (Agnihotry et al., Cochrane Database Syst Rev 2019-05-01, PMID 31145805).
  6. At the visit. Ask which diagnosis it is — irreversible pulpitis, apical periodontitis, or neither — because the treatment fork (root canal, pulpotomy, extraction) and the coverage fork both depend on that sentence (Edwards et al., Br Dent J 2025-04-01, PMID 40217034) (El Karim et al., Int Endod J 2026-06-01, PMID 42370524).
  7. After. If a tooth was removed, the diet, smoking and dry-socket pages below were written for the next seven days; they are where the after-care numbers live.

Frequently asked questions

Why is my tooth more sensitive to hot than to cold? Because inflamed pulp tissue is a pressure chamber and heat increases blood flow into it, while cold transiently reduces it; the experimental record shows inflammation correlates with thermal hyper-response (CXCL8, r = 0.641) and that pulpal pressure itself aggravates pulpitis (Bhat et al., Int Dent J 2026-08-01, PMID 42217380) (Hu et al., Int Dent J 2026-06-01, PMID 42364415). If hot now relieves the pain and cold makes it worse, that is a classic late-pulpitis pattern and an appointment, not a remedy, is the answer.

Can I take ibuprofen and paracetamol together? In the trials, using them together is what produced the biggest drops in pain: diclofenac plus acetaminophen led the network ranking at 6–8 hours (MD −6.28) (Kelidari et al., BMC Oral Health 2026-01-01, PMID 41588508), and a fixed paracetamol-ibuprofen combination matched a 200 mg diclofenac patch through 72 hours in 108 patients (Khanal et al., Ann Med Surg (Lond) 2025-12-01, PMID 41377332). The combination is not licence to exceed either ceiling; the Copenhagen patients show exactly how that goes wrong (Larsen et al., Basic Clin Pharmacol Toxicol 2025-02-01, PMID 39825523). Ulcer disease, kidney disease, blood thinners, late pregnancy and some asthma make nonsteroidal drugs the wrong choice — that is a clinician’s judgement call, not a label one (Gazal et al., Saudi J Anaesth 2026-01-01, PMID 41710634).

How many milligrams of paracetamol is too much? This page will not give you a number, because the only defensible ceiling depends on weight, liver function, alcohol, and what else you have already taken today — and because the study in dental patients found the maximum daily dose was known by 39% of them while 9% exceeded it (Larsen et al., Basic Clin Pharmacol Toxicol 2025-02-01, PMID 39825523). Read the box, count every product that contains the same ingredient, and ask the pharmacist, since 54% of the analgesics in that study came from a place where guidance should have been available.

Is there anything that works instantly? Nothing oral is instant in these records: the fastest measured advantage was the nonsteroidal premedication effect at six hours (Mrini et al., J Clin Exp Dent 2025-12-01, PMID 41716365), and the only “immediate” physical measures with any trial support were applied by a clinician — cold on the injection site (Anjitha et al., J Conserv Dent Endod 2025-12-01, PMID 41438450), cryotherapy alongside a nerve block (Katyal et al., Med Int (Lond) 2025-09-01, PMID 40747149), or cement over an open cavity in one emergency-department case (Chinnock et al., J Am Coll Emerg Physicians Open 2026-04-01, PMID 41808683). A topical benzocaine product may reduce pain for 20 to 30 minutes, at low certainty (Miroshnychenko et al., J Am Dent Assoc 2023-01-01, PMID 36608963).

My child woke up screaming and there is no swelling. What now? Cool the room, an upright carry, and a weight-based dose of whichever analgesic is appropriate for the age — ibuprofen has placebo-controlled data in children with pain intensity measured within two hours (SMD −1.12) (Pessano et al., Cochrane Database Syst Rev 2024-01-01, PMID 38180091) (Miroshnychenko et al., J Am Dent Assoc 2023-05-01, PMID 37105668). Do not add a second nonsteroidal, and do not use an adult’s leftover opioid: among paediatric dental visits involving opioids, 10.1% resulted in an opioid-attributable outcome (Liu et al., J Am Dent Assoc 2025-12-01, PMID 41081662). Then book an examination: the paediatric record describes caries as a cause of sleep disturbance in children, in both directions (Zhang et al., Clin Exp Dent Res 2025-08-01, PMID 40792606).

Do I need antibiotics before the dental appointment? Not for pain. Two Cochrane reviews find no pain or swelling benefit from antibiotics given with a procedure and no evidence at all on antibiotics without a procedure (Agnihotry et al., Cochrane Database Syst Rev 2019-05-01, PMID 31145805) (Cope et al., Cochrane Database Syst Rev 2024-05-01, PMID 38712714). If you have spreading swelling, fever or trouble swallowing, that is an emergency-department presentation — where the antibiotic question is asked by someone who has examined you, with the airway in mind.

Will the emergency department pull the tooth? Sometimes they will do anything to keep you out of the ward: emergency room services provided by a dentist are a covered line in Utah’s Medicaid benefit, while the dental plan itself says no to inpatient hospital services and to nitrous oxide analgesia (data as of 4 February 2026; see Additional documents). The more likely sequence is imaging, analgesia and a referral; the definitive treatment needs the dental chair, which is why the coverage rules put the emergency exam and drainage on the no-authorisation list (Green et al., Am J Emerg Med 2025-03-01, PMID 39764906).

Glossary: kitchen words ↔ chart words

What you say at home What is in the notes How studies measure it
«throbbing, and worse when I lie down» Spontaneous, positional pain; symptomatic irreversible pulpitis until proven otherwise SF-MPQ subscores, pain rating index, associations with posture and sleep (Wang et al., Int Dent J 2025-08-01, PMID 40609202) (Hu et al., Int Dent J 2026-06-01, PMID 42364415)
«hot makes it spike» Thermal hyper-response, exaggerated pain score Gene expression correlated with thermal testing; standardised sensory testing (Bhat et al., Int Dent J 2026-08-01, PMID 42217380)
«the tablets do nothing» Analgesic-resistant pain Rescue-analgesic consumption, 4.6% versus 34.4% in premedication trials (Mrini et al., J Clin Exp Dent 2025-12-01, PMID 41716365)
«my jaw won’t open» Trismus; possible masticator-space involvement Binary swelling outcomes and mouth-opening measures; red flag for hospital assessment (Cope et al., Cochrane Database Syst Rev 2024-05-01, PMID 38712714)
«half my face is numb and tingly» Paraesthesia; consider non-odontogenic causes Diagnostic algorithm with a differential list, not an imaging protocol (Edwards et al., Br Dent J 2025-04-01, PMID 40217034)
«they gave me antibiotics and sent me home» Antibiotic without source control Self-reported prescribing prevalence; 19.2% overall, 26.9% in general practice (Delgado-Giugni et al., Antibiotics (Basel) 2025-12-01, PMID 41594051) (Ambroise et al., Sante Publique 2026-01-01, PMID 42168090)
«can I have something stronger?» Request for opioid escalation Dentist-reported practice challenges, over 80% of respondents (Polk et al., BMC Oral Health 2026-03-01, PMID 41845294)
«it stopped hurting on day four» Necrosis of the pulp; loss of the sensory source Pulpitis versus necrosis cohorts with quality-of-life follow-up (Arifin et al., Odontology 2025-04-01, PMID 39352670) (El Karim et al., Int Endod J 2026-06-01, PMID 42370524)

Related reading on this site: what to do when a tooth chips or breaks, what to eat and drink in the first days, how long to wait before smoking or vaping again and which swelling signs mean the emergency department.

How this page was built, and what it cannot tell you

Method: candidate records were located in Europe PMC with queries matching what people actually type («tooth pain at night what to do», «how to get rid of a toothache fast», «болить зуб вночі що робити», «чим зняти зубний біль в домашніх умовах», «dolor de muelas por la noche», «qué hacer para el dolor de muelas»), and every record used — authors, journal, volume, pages, DOI, PMID, open-access flag, citation count — was re-fetched on the day of publication and is reproduced below with its DOI. Numbers come from the retrieved abstracts, intervals included. Where the closest available evidence comes from a different population (third-molar surgery rather than spontaneous toothache, post-endodontic pain rather than pre-appointment pain, ambulatory children after general anaesthesia rather than a child in a bedroom), the article says so in the same sentence as the number, because the transfer is the weak point. The figure uses one study and one unit. Coverage statements were read from the CMS InsureKidsNow benefit summary for Utah Medicaid (data as of 4 February 2026), and the cost sentence describes a cached extract, not a quote.

What this page cannot tell you: which of eleven painful conditions you have tonight, because that requires a sensibility test, a percussion test and usually a radiograph (Wang et al., Int Dent J 2025-08-01, PMID 40609202) (Edwards et al., Br Dent J 2025-04-01, PMID 40217034); and what dose is safe for your liver, kidneys, stomach, blood pressure, pregnancy or breastfeeding history. What it can tell you is where the reliable benefit sits — a correctly timed non-opioid combination, cold used early, no antibiotic, and an appointment kept (Carrasco-Labra et al., J Am Dent Assoc 2024-02-01, PMID 38325969) (El Karim et al., Int Endod J 2026-06-01, PMID 42370524) — and how quickly the picture changes if swelling, fever or swallowing difficulty appears (Green et al., Am J Emerg Med 2025-03-01, PMID 39764906).

Sources

Peer-reviewed evidence

Bibliographic records above were retrieved and verified programmatically from Europe PMC (authors, journal, volume, pages, DOI, PMID, citation count). Coverage, guideline and regulatory statements are quoted from the cited public documents with their dates; verify against the current version before relying on them. This article is not medical or dental advice.

Exit mobile version