When can I eat and drink after a tooth extraction? What should I eat?

When can I eat and drink after a tooth extraction? What should I eat?

TL;DR

A UK guide, originally for dental patients taking antithrombotic medicines, advises resting until the anaesthetic has worn off and the clot has formed—usually 2 to 3 hours. On the day, avoid hot drinks and hard foods, do not suck forcefully or chew on the treated side, and wait until the next day to rinse [F3]. There is no universal number of days [F24].

This is general health education based on international literature; it does not address the insurance or legal system of any particular country. Care and payment arrangements depend on where you receive care. The clinical evidence here comes from international journals and a Scottish (UK) clinical-guidance patient leaflet (F-Units marked `geo: universal`; eligibility limits are stated in each F-Unit). Only the two passages on the duty to inform and the health-education positioning cite Taiwan law (F-Units marked `geo: TW`); when obtaining care outside Taiwan, those two passages must be read against local requirements.

This is not the same question as “how long does the extraction socket take to heal?”

Another card on this site explains the extraction-socket healing timeline—blood clot, granulation tissue, woven bone, mineralisation, and remodelling—and the definition, red flags, and review criteria for dry socket [F25]. This card does not repeat that material; it cites it only when needed (link in the internal citation chain below).

This card addresses a different question: after you leave the clinic today, when may you drink water, when may you eat, what might you eat or avoid, and how firm is the evidence for each point?

First, what does the evidence for this question look like?

Postoperative diet is frequently searched but unevenly supported by evidence. After searching PubMed from patient questions and obtaining a clinical-guidance leaflet, this site’s assessment is [F24]:

• Direct clinical studies are available for recovery time of soft-tissue sensation after local anaesthesia [F4], risk of self-injury while sensation is absent [F5], day-by-day postoperative symptoms and recovery times [F13][F14][F15][F16], the association between smoking and dry socket [F9][F10], and using a straw [F11].
• Clinical guidance states, but does not itself cite randomised trials for “avoid hot drinks and hard foods on the day,” “do not suck forcefully,” “start rinsing only the next day,” and “do not chew on the treated side” [F3]. These are verifiable words in guidance, not randomised-trial conclusions. The leaflet is for dental patients taking anticoagulants or antiplatelet medicines, not all extraction patients [F3].
• This search found no direct clinical evidence for a causal link between food temperature and post-extraction bleeding, spicy food, ice taken in the mouth, alcohol, or particular supplements accelerating healing. Each is labelled below; this card does not fill gaps with assumptions [F24].

In other words, this card is meant to help you understand the postoperative instructions your clinician gives you and identify questions to ask before leaving; it does not replace those instructions. Under Article 81 of Taiwan’s Medical Care Act, medical institutions must inform the patient or specified related persons about the condition, treatment plan, procedure, medicines, prognosis, and possible adverse reactions [F21]. A medicolegal review also notes that people who are anxious or in pain may not retain oral information, so postoperative instructions should be provided in writing [F20]. You may ask the clinic for a written copy.

Stage 1: the anaesthetic has not yet worn off (the first few postoperative hours)

At this stage, the key issue is not what to eat but whether eating and drinking should wait.

• Soft-tissue anaesthesia lasts longer than many people expect. In a multicentre randomised controlled trial, the control group without a reversal drug had median return of sensation at 155 minutes for the lower lip, 125 minutes for the tongue, and 133 minutes for the upper lip [F4].
• Biting a numb lip, cheek, or tongue without noticing is a recognised type of risk. A review comparing local anaesthetics states that shortening soft-tissue anaesthesia may reduce self-inflicted injuries after an inferior-alveolar nerve block [F5].
• The guidance is direct: rest until the local anaesthetic has worn off and the clot has fully formed, which usually takes 2 to 3 hours [F3].

The two anaesthesia studies were not extraction surgery (they involved restorative or cleaning procedures and a paediatric population). This card cites only the mechanism-level facts that soft-tissue anaesthesia can last several hours and absent sensation can permit self-injury; it does not estimate an individual time [F4][F5]. Follow your operating clinician’s instruction for when you may first drink water and begin eating [F21].

Stage 2: the day of surgery (the blood clot is central)

All dietary limits on the day point to one thing: do not dislodge the newly formed clot.

• Literature describes dry socket as a socket that may be partly or completely without a clot [F6]. Another systematic review describes partial or total post-extraction clot loss causing severe pain that usually begins 1 to 5 days after surgery [F7]. For the full distinction, see the dry-socket section of KM-DENTAL-01; this card does not duplicate it [F25].
• The guidance leaflet gives four instructions for that day: avoid hot liquids and hard foods, avoid forceful sucking or disturbing the clot, avoid chewing on the treated side, and do not rinse before the next day [F3].
• The evidentiary status of those four sentences matters: they are from the March 2022 patient postoperative leaflet of the Scottish Dental Clinical Effectiveness Programme (SDCEP), for dental patients taking anticoagulant or antiplatelet medication; the leaflet does not cite randomised trials supporting each instruction [F3]. This card cites it as verifiable clinical-guidance wording, not as randomised-trial evidence.
• The mouth is not an environment in which a wound can remain still easily. A review describes it as a complex microenvironment affected by oral movements, salivary flow, and bacterial biofilms; these factors can contribute to delayed socket healing and post-extraction complications [F19]. That mechanism helps explain why postoperative instructions contain so many details.

What to do about bleeding: a systematic review included local haemostatic measures such as gauze pressure [F18]. The patient leaflet says that if bleeding restarts after leaving the clinic, press on the area for 30 minutes with a folded, clean, damp handkerchief or gauze pad, and contact the dentist if it does not stop [F3]. The same systematic review states that interrupting antithrombotic medicines because of fear of bleeding, or after postoperative bleeding, increases the risk of thromboembolic events [F18]. If you take an anticoagulant or antiplatelet medicine, report the bleeding to your clinician; do not stop or change the dose on your own.

Stage 3: the first 3 postoperative days (the symptom plateau)

• A patient-reported study of 339 people undergoing five types of alveolar surgery used a 10 cm visual analogue scale to record bleeding, swelling, pain, and bruising daily for postoperative 1 week and at day 14. In the first 3 healing days, transalveolar extraction had a higher area under the curve for overall bleeding and pain than the other groups (mean bleeding 5.6; mean pain 7.5) [F8].
• A quality-of-life study of mandibular third-molar surgery (92 patients, including eating and speaking ability) concluded that lower third-molar surgery significantly affects quality of life and environment, particularly in the first 3 days after extraction [F14].
• A prospective longitudinal study of upper third molars (47 included in analysis) found pain fell linearly over 7 days, with statistically significant relief between day 2 and day 3; upper third-molar removal significantly affected quality of life, particularly in the first 2 days [F15].
• Taken together, these three studies support only this: the first 2 to 3 days are a period when discomfort is more concentrated and quality of life more noticeably affected, then it declines. This is a population-level pattern, not an inevitable individual outcome [F8][F14][F15]. The two quality-of-life questionnaires included eating ability, but their abstracts do not report a numeric eating subscore, so this card does not turn them into “how many days until normal eating” [F14][F15][F24].

What should you eat in this period? After the day of treatment, the guidance no longer sets foods by day; it shifts to cleaning: beginning the day after treatment, gently rinse with warm salt water (one teaspoon of salt in one glass of water) 3 to 4 times daily for 5 days [F3]. The evidence level for warm salt water needs an honest description: a systematic review and meta-analysis of 8 randomised studies found no significant difference in dry-socket incidence between warm saline mouth baths and other antimicrobial rinses (`P > 0.05`), and the authors noted high risk of bias in most included studies [F17]. “No significant difference” means it was not shown to be better or worse than other rinses; it does not mean “useless” or “do not do it.” Whether and how to rinse should follow your postoperative instruction [F3][F21].

Stage 4: after 1 week (when does “normal eating” return?)

• A prospective clinical trial enrolled 740 people having all four third molars removed and obtained recovery data from 630. Patients completed a daily quality-of-life measure for 14 postoperative days, covering pain, daily activities, oral function, and other symptoms. Most quality-of-life measures recovered within 5 days, but recovery of pain to “little or none” lagged behind the other measures; 22% of patients were treated for delayed healing [F13].
• Recovery speed differs between populations. A study of 958 subjects enrolled at 9 academic centres and 12 community practices defined recovery as the days needed for patients to report no or only a little trouble. Apart from mouth opening, people aged 21 years or older had significantly slower recovery on all quality-of-life items than younger people, and women had significantly longer recovery times than men for all outcomes [F16].
• In plain language: “about 5 days” is a population-level observation in that study, not a warranty period. The source says most quality-of-life measures recovered within 5 days; it does not provide a median or mean [F13]. Age, sex, and surgical difficulty can change your actual pace [F13][F16].

What to eat: this card gives three criteria, not a food list

Most online lists of “20 foods you can eat after an extraction” lack clinical-study support. This card can cite only two nutrition-related items, and both need clear limits:

• Protein intake and early wound healing: a pilot study of 25 patients used a food-frequency questionnaire and 24-hour recall to assess habitual protein intake in the preceding month, then assessed early wound-healing index after impacted mandibular third-molar surgery. Protein intake was positively associated with the day-7 early wound-healing index (`p = 0.019`) and lower day-1 pain scores (`p = 0.009`). In the same study, however, the high-protein group had a greater decrease in mouth opening on day 3 (`p = 0.004`), more facial swelling on day 3 (`p = 0.023`), and no significant differences in OHIP-14 scores at any time point [F12]. This is a pilot study of 25 people measuring what they had habitually eaten, not a trial of eating more protein after surgery; the authors themselves call for larger studies [F12].
• Malnutrition and delayed healing: mice fed a low-casein diet for two weeks before extraction showed delayed wound healing on day 7 with higher inflammation-related indicators [F23]. An animal experiment cannot be extrapolated to people. It is cited only to show that being unable to eat or eating too little is not mechanistically irrelevant [F23].

Accordingly, this card gives criteria rather than a list. The following three are this site’s communication framework derived from F3 and F19, not clinical guidance [F2]:

1. No forceful biting—the guidance says to avoid hard food and chewing on the treated side on the day [F3]; oral movement itself is listed as one factor that may delay socket healing [F19].
2. Do not burn your mouth—the guidance says to avoid hot liquids on the day [F3]. Important: this site found no direct clinical trial on food temperature and post-extraction bleeding, so this is guidance wording, not a trial conclusion [F3][F24].
3. No forceful sucking needed—the guidance says to avoid sucking hard or disturbing the clot [F3]. Straw use itself has a direct study whose conclusion may differ from what you have heard (next section) [F11].

Being able to eat and getting enough to eat are part of postoperative care. If pain or limited opening means you can hardly eat, that is something to report at a review visit, not something you should simply endure [F12][F13][F22].

What to avoid, and why: evidence level item by item

Smoking: the most densely cited item in this card

• A systematic review of 11 studies found that smokers had more than 3 times the odds of dry socket; combined incidence was about 13.2% in smokers and about 3.8% in non-smokers [F7].
• An evidence-based dentistry commentary on that review added that the 11 studies came from 10 countries and included 10,195 patients (3,007 smokers and 7,188 non-smokers); every included study was only at the third or fourth level of a five-level evidence scale [F10].
• One honest qualification: the guidance leaflet cited in this card does not mention smoking. The smoking evidence above comes entirely from a systematic review and its commentary, not from guidance wording [F3][F7][F10]. There is no single standard answer for how long or how to stop; follow your clinician’s plan.

Forceful sucking and straws: one direct study points against the usual claim

• In a study of 60 patients having all four third molars removed, half used a straw with every meal for the first 2 postoperative days. A total of 220 teeth were removed; no dry socket occurred in the maxilla, while 17 occurred in the mandible—8 among straw users (15%) and 9 among non-users (15%). The authors concluded that no evidence showed an increased dry-socket incidence from using a straw in the first 2 days after third-molar extraction [F11].
• Read that carefully: it is one study, with 60 patients, published in a state dental-association journal. The abstract says “randomly selected patients,” but does not describe random allocation or blinding, and does not clearly define the percentage denominator [F11].
• This card’s position is therefore: “straws cause dry socket” currently lacks supporting clinical evidence, but that does not mean you may disregard your postoperative instructions. The guidance still says to avoid forceful sucking or disturbing the clot [F3][F11]. The two statements do not conflict: the first concerns the device, the second the action.

Hot drinks and hot food

• What can be cited is the guidance wording: avoid hot liquids and hard foods on the day [F3].
• This search found no clinical trial of food temperature and post-extraction bleeding or dry socket [F24]. The only mechanism-level support available is impaired temperature and pain judgement while anaesthesia has not worn off [F4][F5].
• This card therefore gives no number of days after which hot food may be eaten. Any fixed-day claim exceeds what the available literature can support [F24].

Alcohol

• This search found no direct clinical study of alcohol use after extraction and dry socket or wound healing [F24].
• Whether alcohol may be used while taking postoperative medicines is an individual medication question; follow your clinician’s advice [F21]. This card gives no alcohol recommendation.

Timing of rinsing

• The guidance says not to rinse before the next day; beginning the day after treatment, gently rinse with warm salt water 3 to 4 times daily for 5 days [F3].
• Warm salt water and other antimicrobial mouth rinses showed no significant difference in dry-socket incidence, and the included studies had high risk of bias [F17].
• This site found no clinical trial of rinse force and clot loss. What can be cited is that literature lists a socket partly or totally without a clot as a descriptive feature of dry socket [F6][F24].

Wisdom-tooth version: why “normal eating” may come later

“When can I drink water after wisdom-tooth removal?” and “when can I eat normally after wisdom-tooth removal?” are often asked separately and deserve a direct comparison.

• The experience in the first days is indeed different: in the patient-reported study, transalveolar extraction had a higher area under the curve for overall bleeding and pain during the first 3 healing days than the other groups [F8].
• Pain is not the only limit on eating; reduced mouth opening is another: a Cochrane review of mandibular wisdom-tooth surgery says surgery is commonly associated with short-term postoperative pain, swelling, and trismus. Less commonly, infection, dry socket, and trigeminal nerve injury may occur [F22].
• There is a citable scale for recovery time: after removal of all four third molars, most quality-of-life measures recovered within 5 days, pain recovery lagged behind, and 22% were treated for delayed healing [F13]; recovery was significantly longer for people aged 21 years or older and for women [F16].
• The dry-socket risk range also differs: the Cochrane review cites a prevalence of 1% to 5% for routine extractions and upwards of 30% for surgically extracted third molars [F9]. A meta-analysis of 28 studies covering 41,859 impacted mandibular third-molar extractions estimated overall prevalence at 6.7% (95% confidence interval 4.6% to 9.1%) and reported considerable heterogeneity [F26].
• But the operative diet principles are not separately set by tooth position in the guidance: the leaflet applies to “dental treatment that is likely to cause bleeding, for example an extraction,” and does not give separate advice by tooth position [F3]. The difference is intensity and duration, not the principle.

Common claims checked one by one

The following records what this site obtained by searching PubMed E-utilities on 2026-08-05. “No direct evidence obtained” means this search did not find a citable clinical study; it does not mean the claim has been disproved [F24].

1. “A straw sucks out the clot and causes dry socket.” One direct study recorded 8 straw users and 9 non-users, both shown as 15% in the source; the authors found no evidence of higher incidence [F11]. Its small sample and limited reporting of method details and percentage denominators do not overturn the guidance to avoid forceful sucking [F3][F11].
2. “You cannot drink anything hot on the day of extraction.” The guidance does say this [F3], but this search found no clinical trial of food temperature and postoperative bleeding [F24]. This is guidance wording, not a trial conclusion.
3. “Eating ice after an extraction reduces swelling better.” This search found no clinical study of ice taken in the mouth after extraction [F24]. External ice application is a different matter; its evidence is in the relevant section of KM-DENTAL-20 [F25].
4. “You cannot eat spicy food after an extraction.” This search found no clinical study of spicy food and extraction wounds [F24].
5. “Taking more supplements makes the socket heal faster.” No clinical evidence was obtained that a particular food accelerates human extraction-socket healing [F24]. The only citable nutrition evidence is the pilot study of habitual protein intake [F12] and the mouse malnutrition experiment [F23]; neither endorses supplements.
6. “If it no longer hurts, it is healed and I can eat normally.” Pain resolution and tissue healing are different. In the third-molar study, pain recovery even lagged behind other quality-of-life measures, and 22% were treated for delayed healing [F13]. See KM-DENTAL-01 for the socket-healing timeline [F25].

<!-- 風險因素 -->

Risk factors: whose postoperative diet needs more caution?

• People who smoke: dry-socket odds were more than 3 times higher, with combined incidence 13.2% versus 3.8%; however, the included studies were third- or fourth-level evidence [F7][F10].
• People taking antithrombotic medicines: postoperative bleeding needs clinician management; stopping antithrombotic medicines on one’s own increases thromboembolic-event risk [F18]. The guidance leaflet cited here was written for this population; its medication passages need individual explanation from your clinician [F3].
• People taking antiresorptive or antiangiogenic medicines, or in an immunosuppressed state: a systematic review of 23 papers found that leading complications in oral surgery were haemorrhage in anticoagulated patients, infection in immunosuppressed patients, and osteonecrosis in people taking antiresorptive drugs [F22].
• People having surgical extraction (flap elevation, bone removal, tooth sectioning): bleeding and pain area under the curve were higher in the first 3 days [F8], and dry-socket risk range was higher [F9][F26].
• Older people and women: recovery time for postoperative quality-of-life measures was significantly longer [F16]. This is a population-level observation, not an individual prediction.
• People who were not eating enough already: the positive association between habitual protein intake and day-7 early wound-healing index is from a pilot study [F12]; evidence that malnutrition delays extraction-wound healing comes from a mouse experiment [F23].

Tooth extraction and aftercare have risks and contraindications. Whether any item above applies to you must be assessed by a dentist for your individual circumstances.

Before the appointment / before leaving the clinic: 7 questions to ask

1. Was mine a simple extraction, or a surgical extraction with a flap, bone removal, or tooth sectioning? How does recovery differ? [F8][F22]
2. About how long will the anaesthetic last? May I drink water before it wears off? When may I start eating? [F3][F4][F5]
3. What may I eat or avoid from today to tomorrow? Can I take written postoperative instructions home? [F3][F20][F21]
4. When should I start rinsing? Should I use warm salt water, how should I rinse, and for how many days? [F3][F17]
5. I take anticoagulants, antiplatelet medicines, or other chronic-disease medicines. What should I note about postoperative bleeding and diet? May I drink alcohol? [F18][F21]
6. I smoke. How should I understand my dry-socket risk for this extraction? Can the clinic help refer me to smoking-cessation resources? [F7][F10]
7. What signs need an immediate review (bleeding that will not stop, pain that worsens rather than improves, being barely able to eat), and whom do I contact outside clinic hours? [F3][F13][F18]

<!-- 常見問題 -->

How to find a clinic that provides this service

This site lists no clinics and does not recommend, rate or compare any medical institution.

To find a clinic that provides this service, use the following official lookup channels yourself —

they are maintained by the competent authorities and are more current than any list:

>

- The Ministry of Health and Welfare's medical-institution lookup: search by city or county,

township and specialty (dentistry) for institutions that have completed practice registration.

The results show the institution's name, address and registered specialty. This is the first place

to confirm whether a clinic is lawfully registered.

- The National Health Insurance Administration's contracted-institution lookup: check whether a

clinic is under contract with National Health Insurance. This determines which items are covered

and which are self-paid, and therefore what kind of fee explanation you will be given.

>

Both systems are on the respective authorities' official websites; searching for the

medical-institution lookup or the contracted-institution lookup will find them. This site does not

reproduce the URLs, because they change; use whichever entry point the authority currently publishes.

>

In an emergency, do not spend time looking up lists. If your breathing, swallowing or speech is

affected, or swelling is spreading, go to the nearest emergency department.

Chinese labels to look for on the official pages

These official systems, datasets and files are published in Chinese only. The English names above are this site's renderings; on the page itself you will see the following strings, so search for these:

>

- 「醫事機構查詢」

- 「特約醫事機構查詢」

- 「特約院所查詢」

Compliance note

This is health education information under Article 87 of Taiwan’s Medical Care Act, not medical advertising, and it does not recommend a particular clinic [F0]. Tooth extraction and post-extraction care have risks and contraindications. Actual treatment methods and outcomes vary by person and require a dentist’s assessment. The stage framework and dietary criteria in this card are for communication during care; they cannot replace postoperative instructions and clinical diagnosis from your treating dentist.

Every clinical statement in this article is mapped line by line to its cited source (see the sources and evidence chain below). It has not been clinically reviewed by a licensed practitioner. This is health information, not individual advice; assessment by a clinician is required.