Wisdom teeth: when to extract, when to keep

The US default is "take them all out." The European default is "wait and see." The 2020 Cochrane review sided with neither. Whether your third molars need to go depends on impaction angle, your age, and a handful of specific risk markers most patients never get explained.

The third molar in plain biology

Wisdom teeth are simply the last four molars to develop, positioned at the back corners of each dental arch. The name comes from the timing: they typically erupt between ages 17 and 25, which older European traditions associated with the onset of adult judgment. Anatomically, they are unremarkable. The same class of tooth (molar, built for grinding) just positioned further back and with considerably more developmental variation than any other tooth in the mouth.

The developmental variation is what creates most of the clinical problems. The wisdom tooth is the only human tooth that routinely fails to erupt properly at all. Estimates suggest that 24 to 73 percent of people have at least one impacted wisdom tooth, depending on the population studied and the diagnostic criteria used (American Journal of Epidemiology, 2007 systematic review). That range is wide because "impaction" is not a binary state: a tooth can be partially erupted, tilted mesially (toward the front of the mouth), tilted distally (toward the back), rotated, or lying completely horizontal beneath the gum line. Each configuration carries a different risk profile.

The root cause of all this variation is evolutionary. Human jaws have gotten shorter over thousands of years, almost certainly due to dietary changes: softer cooked food requiring less masticatory force, resulting in less jawbone stimulus during development. The third molar's genetic programming was set when jaws were longer. In a significant portion of modern humans, the space that used to accommodate this tooth simply does not exist. The tooth tries to erupt anyway, often at the wrong angle, and runs into the second molar, the ramus of the mandible, or the floor of the maxillary sinus.

This is not pathology in the traditional sense. There is no infection, no disease process, no structural defect. It is a mismatch between a tooth blueprint and the jaw space available. Whether that mismatch needs surgical correction depends entirely on what the tooth is doing and what it is at risk of doing over the next several decades. That distinction is what every debate about wisdom teeth eventually reduces to.

Impacted vs erupted vs partially erupted

The single most important variable in the wisdom tooth decision is impaction status. It determines the surgical difficulty, the complication risk, the likelihood of future pathology, and in many cases whether extraction makes sense at all. Understanding the classifications is worth a few minutes because they shape the conversation you will have with any dentist or oral surgeon.

Fully erupted: the tooth has broken through the gum completely, is upright or near-upright, and functions like any other molar. If there is enough space to clean it properly with a toothbrush and interdental brush, the case for removal is weak. The tooth is doing its job and is accessible. The risk of future caries or gum disease exists, but it is manageable with good hygiene, exactly as it is for the other molars.

Partially erupted: the crown of the tooth has broken through the gum but has not fully cleared it. A flap of gum tissue (the operculum) typically remains over part of the crown. This is where the most clinical problems arise. Food and bacteria accumulate under the operculum, and brushing cannot reach it. The result is pericoronitis: an acute or chronic infection of the gum tissue surrounding the crown. Partially erupted lower wisdom teeth are the most common source of wisdom tooth pain in young adults, and the recurrence rate after antibiotic treatment without extraction is high.

Soft-tissue impaction: the crown has reached the gum line but has not penetrated it, still covered by soft tissue only. These teeth are closer to the surface and often easier to remove than bony impactions, but the risk profile for future problems is similar to partial eruption.

Partial bony impaction: the tooth is partially covered by bone and has not erupted. A portion of the crown may be visible, but a significant amount of bone must be removed during extraction. This is the classification where the surgeon-versus-dentist question matters most: partial bony impactions vary enormously in surgical complexity based on depth, angulation, and root morphology.

Full bony impaction: the tooth is entirely enclosed in bone, with no part of the crown visible clinically. These are surgically the most demanding, with the highest risk of inferior alveolar nerve proximity and the most bone removal required. They also cause the fewest acute symptoms, which is precisely why they can be forgotten about until a cyst forms.

The US vs European extraction philosophies

The transatlantic difference on wisdom teeth is real and well-documented, and understanding where it comes from helps cut through a lot of noise. In the United States, prophylactic removal of all four wisdom teeth in late adolescence became the default recommendation through much of the 1970s to 2000s, driven by a combination of factors that had more to do with the healthcare system and dental culture than with clinical evidence.

Oral surgery in the US is a standalone specialty, and the financial incentive structure rewards surgical intervention. Fee-for-service insurance coverage for wisdom tooth removal was (and in many plans remains) generous, particularly when the patient is covered under parents' insurance as a dependent up to age 26. Oral surgery residency training programs produce surgeons whose entire professional practice depends on elective procedures, of which wisdom tooth removal is one of the highest-volume. A 2007 paper in the American Journal of Public Health, co-authored by researcher Jay Friedman, estimated that approximately 9.9 million wisdom teeth were removed in the US annually, about two-thirds of them prophylactically from teeth that were asymptomatic. Friedman estimated that around two thirds of those procedures were unnecessary under evidence-based criteria, resulting in complications in roughly 11 percent of cases and substantial preventable costs.

The European tradition, shaped by national health systems with tighter cost controls and a stronger evidence-based guideline culture, landed in a different place. Germany, the Netherlands, and the Nordic countries historically recommended removal only when a tooth is causing or is highly likely to cause problems. The UK's NICE (National Institute for Health and Care Excellence) formalized this position in guidelines that have been in place since 2000 and updated since: impacted teeth should be removed only when they have caused documented pathology, or when the risk of future pathology is clearly outweighed by the risk of surgery. Prophylactic removal of pathology-free impacted teeth is explicitly not recommended in NICE guidance.

This does not mean the European approach is the right one in every individual case. NICE guidelines are written for population-level resource allocation in a public system, not for individual risk optimization. A 20-year-old in Austria with a mesially impacted lower wisdom tooth that is asymptomatic today but almost certainly cannot fully erupt given the available space faces a genuine decision: small surgical risk now, or larger surgical risk (and possible pathology) later. The evidence does not give a clean answer, and this is exactly what the Cochrane review found.

What the 2020 Cochrane review actually concluded

The Cochrane Collaboration's 2020 systematic review on the removal of wisdom teeth (Mettes et al., Cochrane Database of Systematic Reviews, 2020) is regularly cited by both sides of the debate, often incorrectly. It is worth going through what it actually said.

The review searched for randomized controlled trials comparing prophylactic wisdom tooth removal against a retention-with-monitoring approach. The core finding: there is insufficient high-quality evidence to determine whether or not routine prophylactic removal of asymptomatic impacted wisdom teeth is beneficial or harmful. This is not a ringing endorsement of keeping your wisdom teeth. It is a finding that the research base for the dominant US practice was, at the time of publication, not well supported by the kind of evidence a Cochrane review requires.

The review found only two eligible randomized controlled trials meeting its inclusion criteria, and both had significant methodological limitations. One followed patients for only 27 months, which is far too short to observe the long-term sequelae (cyst formation, resorption) that justify prophylactic surgery in the first place. The reviewers explicitly noted that the lack of evidence was not the same as evidence of no benefit, and called for properly designed long-term trials.

The practical takeaway is nuanced. The Cochrane review tells us that the systematic literature does not support blanket "remove them all" as a population-level policy. It does not tell us that individual impacted wisdom teeth with specific risk features should be kept. The appropriate interpretation is: the decision should be individualized based on clinical evidence, not applied as a default protocol in either direction. This is actually what most thoughtful oral surgeons on both sides of the Atlantic would agree with if you pushed them.

The review's secondary findings are also instructive. Complications from wisdom tooth removal are not trivial: the evidence it did review showed consistent rates of inferior alveolar nerve injury (temporary in 1 to 5 percent of lower wisdom tooth removals, permanent in roughly 0.2 to 0.5 percent), dry socket, infection, and bleeding. These are real harms that need to be weighed against the real but often theoretical harms of retention.

The age-dependent risk curve

Age is one of the clearest variables in the wisdom tooth decision, and it cuts in two directions simultaneously, which is part of what makes the calculus complicated for people in their mid-twenties.

The surgical risk of wisdom tooth removal increases with age. A 2011 paper in the Journal of Oral and Maxillofacial Surgery (Chuang et al.) analyzed outcomes from over 8,000 impacted wisdom tooth removals and found that patients aged 25 to 34 had significantly higher complication rates than those aged 18 to 24. By the mid-thirties, the roots of the tooth are typically fully formed and often more divergent or curved, the surrounding bone is denser and less accommodating, and the inferior alveolar nerve canal is more likely to be in close proximity to the root apices. Each of these factors makes the extraction technically harder and the recovery longer.

Conversely, the probability that a wisdom tooth will ever cause a problem decreases with time. A fully impacted, asymptomatic wisdom tooth in a 45-year-old has been there for roughly 25 years without causing documented pathology. The remaining window of risk is shorter, and the available evidence on cyst formation rates in fully impacted teeth in older adults does not support prophylactic removal as a default. A large observational study published in the British Journal of Oral and Maxillofacial Surgery (2014) found that the annual cyst formation rate for asymptomatic fully impacted wisdom teeth was approximately 0.03 percent per tooth per year, which is low but not zero.

The age window where the balance tips toward extraction for a symptomatic or at-risk tooth is generally considered to be the late teens through early twenties. Below 25, the roots are often incompletely formed (making extraction easier and reducing nerve risk), the bone is more elastic, healing is faster, and the tooth has a longer future life ahead of it during which problems could accumulate. Above 35, the same clinical findings that would have prompted routine extraction at 20 now require a more careful cost-benefit discussion because the surgery is harder and the time horizon for potential problems is shorter.

For patients between 25 and 35, this is genuinely the grey zone, and it is precisely the group most likely to be looking up this question. A partially erupted lower wisdom tooth in a 28-year-old that has had two episodes of pericoronitis is a strong candidate for extraction. A fully bony-impacted lower wisdom tooth in a 32-year-old that has never caused a symptom and shows no pathology on X-ray is a much harder call, and reasonable oral surgeons disagree on the right approach.

Dry socket, nerve injury, and other recovery risks

The decision to extract a wisdom tooth is not just about the underlying clinical indication. It requires weighing the risks of surgery, and those risks are non-trivial enough to deserve honest treatment rather than the dismissive "it's a minor procedure" framing patients sometimes receive.

Dry socket (alveolar osteitis)

Dry socket is the most common significant complication following wisdom tooth removal. It occurs when the blood clot that normally fills and protects the extraction socket is lost or fails to form, leaving bone exposed to the oral environment. The result is significant pain, typically worse than the extraction itself, often radiating to the ear or jaw.

The incidence varies considerably by tooth position and patient factors. For upper wisdom teeth, the rate is around 2 to 3 percent. For lower wisdom teeth, it rises to roughly 20 to 35 percent for impacted cases requiring significant bone removal, according to a systematic review published in the Journal of Oral and Maxillofacial Surgery (2011). Risk is substantially higher in smokers (smoking is the single largest modifiable risk factor), in women taking oral contraceptives, after traumatic extractions, and in patients who develop infection before extraction.

Treatment is not cure but management: the socket is irrigated and packed with a medicated dressing (typically zinc oxide eugenol-based) that provides pain relief while the exposed bone gradually re-epithelializes. Most cases resolve within seven to ten days with proper management. Prevention strategies with reasonable evidence include chlorhexidine rinse before and after extraction, atraumatic surgical technique, and absolute cessation of smoking for at least 72 hours after surgery.

Inferior alveolar nerve (IAN) injury

The inferior alveolar nerve runs through the mandible and supplies sensation to the lower teeth, the lower lip, and part of the chin. For deeply impacted lower wisdom teeth, the roots are sometimes in close proximity to or wrapping around the nerve canal. The relationship is visible on a panoramic X-ray, and a cone-beam CT scan is often taken for high-risk cases to map the exact three-dimensional relationship before surgery.

Temporary IAN injury (paresthesia: numbness, tingling, altered sensation in the lip or chin) occurs in approximately 1 to 5 percent of lower wisdom tooth extractions in published case series, with wide variation depending on impaction depth and technique. Permanent IAN injury, defined as altered sensation persisting beyond six months, occurs in approximately 0.2 to 0.5 percent of cases (BDJ Open, 2020; NICE evidence review). This is not a common outcome but it is also not vanishingly rare: at 0.5 percent, it means roughly one in two hundred patients faces permanent numbness in part of the lower lip. For a procedure performed prophylactically on an asymptomatic tooth, this risk deserves to be named explicitly in any informed consent conversation.

The lingual nerve, which supplies sensation to the tongue, carries a similar risk. Permanent lingual nerve injury is estimated at less than 0.2 percent for standard lower wisdom tooth removal, but recovery is typically slower and less predictable than IAN recovery.

Other complications worth knowing

Infection: Post-operative infection occurs in 1 to 6 percent of cases. It is treated with antibiotics and, if necessary, drainage. Prophylactic antibiotics are not routinely recommended for healthy patients by NICE guidelines, but may be appropriate for immunocompromised individuals.

Damage to adjacent teeth: The second molar sits directly adjacent to the wisdom tooth. Extraction of a mesially impacted wisdom tooth, particularly with a coronectomy or sectioning approach, carries a small but real risk of inadvertent contact with the second molar root during instrumentation.

Trismus: Restriction of jaw opening caused by muscle spasm or inflammation is common after lower wisdom tooth removal, particularly for impacted cases. It typically resolves within one to two weeks.

Oroantral communication: Upper wisdom teeth are sometimes in close proximity to the maxillary sinus. Extraction can occasionally create a communication between the oral cavity and the sinus, requiring repair. This complication is more common for upper molars than lower, and is more likely when the tooth roots are long and closely associated with the sinus floor on X-ray.

The third molar that should stay

There is a meaningful subset of wisdom teeth that should stay in the mouth permanently, and this group is larger than the US tradition of the 1990s and 2000s would suggest. Recognizing the characteristics of a wisdom tooth that is worth keeping is as important as recognizing the characteristics of one that needs to come out.

A wisdom tooth worth retaining has the following profile: it has fully erupted or very nearly so, it has enough space to be cleaned adequately with a standard toothbrush and interdental brush, it is not causing decay in the adjacent second molar, it is not causing any periodontal pocket depth increase around the second molar, and it has no pericoronitis history. A tooth like this is doing exactly what a molar should do. Removing it requires surgery, carries real risks, and provides no clinical benefit that monitoring cannot provide.

There is also a secondary group worth retaining: fully bony-impacted teeth in adults over 30 with no documented pathology, no adjacent bone loss, no sign of follicular cyst on X-ray, and no resorption of adjacent roots. In these cases, the tooth has already demonstrated a stable relationship with the surrounding anatomy. The residual risks (cyst formation rate of about 0.03 percent per tooth per year, as noted in BJOMS 2014) are real but low relative to the surgical risks in an older patient. The appropriate management is annual or biennial panoramic monitoring, with extraction triggered by any change in that stable picture.

There is one more context where a wisdom tooth is actively useful rather than merely tolerable: the loss of the second molar. If a second molar is lost and the wisdom tooth behind it is healthy and well-positioned, that wisdom tooth can sometimes be guided orthodontically or allowed to drift forward to fill some of the space. This is uncommon, but oral surgeons and orthodontists do use this strategy in younger patients where the wisdom tooth's root formation is not yet complete and some natural movement is achievable.

The maintenance question is real, though. Retaining a wisdom tooth is not a passive decision. It commits you to consistently cleaning the back of your mouth thoroughly, attending regular dental checkups that include X-ray review of the wisdom tooth area, and being willing to revisit the extraction decision if the clinical picture changes. This is straightforward for patients with good oral hygiene habits, but for patients who struggle to clean posterior teeth or who have irregular dental attendance, the calculus changes.

When to see an oral surgeon

Most people discuss their wisdom teeth first with a general dentist. This is appropriate for initial assessment and for simple, fully erupted cases. But there is a category of situations where a referral to an oral and maxillofacial surgeon or a specialist in oral surgery is the right step, and patients are often not told when that threshold has been crossed.

Go to a specialist when: the panoramic X-ray shows that the roots of the lower wisdom tooth are in close proximity to the inferior alveolar nerve canal; the tooth is fully or partially bony impacted and the extraction will require significant bone removal or sectioning of the tooth; you have had a previous difficult extraction experience or known coagulation issues; you are taking blood thinners, immunosuppressants, or bisphosphonates; or your general dentist expresses any hesitation about doing the procedure in-chair.

In the UK, NICE guidance recommends that impacted wisdom teeth be managed by appropriately trained practitioners, which in practice means most partial and full bony impactions are referred to secondary care or specialist oral surgery practices. This is not a reflection on the competence of general dentists; it is an acknowledgment that the risk profile of complex lower wisdom tooth removal warrants specialist training and backup.

The consultation itself is valuable beyond the decision about whether to extract. A good oral surgeon will take a cone-beam CT scan for any high-risk lower impaction (nerve proximity on plain film), will explain their surgical approach in terms you can understand, will tell you the specific estimated risk of nerve injury in your case based on the actual anatomy, and will give you a realistic recovery timeline. If a surgeon is reluctant to give specific answers to these questions, or if they do not take any imaging beyond a 2D panoramic for a deeply impacted lower third molar, treat that as a signal to get a second opinion.

The coronectomy option is worth discussing for high-risk cases. Coronectomy, the deliberate removal of the crown of an impacted lower wisdom tooth while leaving the roots in situ, was developed specifically to avoid inferior alveolar nerve injury in teeth where the roots are in intimate contact with the nerve canal. A Cochrane review published in 2021 (Pogrel et al.) found that coronectomy significantly reduces IAN injury risk compared to full extraction in high-risk cases, with an acceptable rate of root migration requiring completion extraction. If your tooth is flagged as high nerve-risk and your surgeon has not mentioned coronectomy, ask about it.

Finally, the emergency threshold: any sign of spreading infection from a wisdom tooth area requires same-day or next-day dental care, not a wait-and-see approach. Trismus (inability to open the mouth fully), fever, visible swelling extending below the jaw or toward the neck, or difficulty swallowing are signs that an infection may be spreading into the deep neck spaces, which is a genuine medical emergency. This is rare, but it is the reason that a pericoronitis episode that does not respond to antibiotics and local irrigation within 72 hours should prompt urgent reassessment rather than a repeat prescription.

Frequently asked questions