A Mallampati score describes how much of the back of the mouth a clinician can see during a standardized exam. A higher class can be one sign of a crowded oral airway, but it does not show what happens while you sleep. It cannot diagnose obstructive sleep apnea, rule it out, measure its severity, locate the site of collapse, or determine treatment.
The score began as an anesthesia airway assessment. Sleep clinicians may record it alongside symptoms, health history, body measurements, and other examination findings. Those two uses overlap, but neither turns the Mallampati score into a stand-alone test 12.
What the Mallampati score measures
The examiner looks at the relationship between the tongue, palate, uvula, and other structures at the back of the mouth. The original Mallampati system had three classes and was developed to help anticipate a limited view during direct laryngoscopy. Samsoon and Young later added a fourth class, creating the four-class modified Mallampati system commonly used today 2.
The exam describes a visible awake anatomy. It does not directly measure:
- airflow during sleep
- the frequency of apneas or hypopneas
- oxygen levels
- sleep stages or arousals
- respiratory effort
- the exact site or pattern of airway collapse
These limits matter because OSA develops during sleep, when muscle activity, sleep stage, body position, and several other factors affect airway stability. A mouth that looks crowded while someone is awake may not produce a predictable number of breathing events. A less crowded view does not guarantee that the airway stays open 1.
How clinicians perform the exam
For the usual modified Mallampati exam, the person sits upright with the head in a neutral position, opens the mouth as widely as possible, and protrudes the tongue as far as possible. The clinician observes the view without asking the person to say "ah" 23.
Phonation changes the result. In a study of 64 surgical patients assessed with and without phonation, saying "ah" systematically improved the visible view. A second group showed a smaller posture effect, and differences between the two observers were substantial 4. That is why a grade recorded after phonation should not be treated as interchangeable with one recorded without it.
Mouth opening, tongue effort, head position, breathing pattern, lighting, and the examiner's interpretation can also affect the class 24. Pain, limited movement, or swelling in the mouth or throat may change either the anatomy or the person's ability to perform the maneuver. A score from an old record may therefore not describe the airway in a new clinical situation 5.
Why a mirror or photo is not a reliable self-test
A mirror, phone photo, or app can use a different camera angle, tongue position, head position, lighting, or breathing pattern from the clinical exam. It can also be difficult to tell whether the entire uvula or only its base is visible. Even a technically correct class would still be only one examination finding, not a diagnosis.
Do not use a self-score to decide whether to seek a sleep evaluation, cancel one, choose a treatment, or judge the safety of anesthesia. If the class matters to a clinical decision, let the clinician performing that evaluation document it.
Mallampati classes 1 to 4
Accepted descriptions vary slightly, especially in how much of the uvula is visible in classes II and III. The consensus description below uses the palate and uvula as the main landmarks. Class IV consistently means that only the hard palate is visible 1.
| Class | Structures visible in a typical modified Mallampati description | What the class does not mean |
|---|---|---|
| I | Complete soft palate and uvula; the fauces and pillars are commonly visible | It does not rule out OSA or guarantee easy airway management |
| II | Complete soft palate and part of the uvula; some accepted descriptions show more of the uvula but not the pillars | It is not a fixed "moderate risk" category |
| III | Base of the uvula, with a limited soft-palate view | It does not diagnose sleeping obstruction or predict its severity |
| IV | Only the hard palate; the soft palate is not visible | It does not prove severe OSA or guarantee difficult intubation |
Some records use Roman numerals and others use Arabic numbers, so "Mallampati IV" and "Mallampati 4" usually refer to the same class.
Mallampati class is not Friedman tongue position
The terminology in sleep medicine is inconsistent. In the modified Mallampati exam described above, the tongue is protruded. In the Friedman tongue position exam, the person keeps the tongue inside the mouth. Friedman tongue position was developed for OSA-related assessment and surgical staging, not as the same maneuver under a second name 21.
When reading a study or medical record, check what the examiner actually asked the person to do. A label alone may not reveal which technique was used.
What a higher score says about sleep apnea risk
Across groups, higher Mallampati classes are associated with higher odds of OSA or a higher apnea-hypopnea index in some studies. A 2025 systematic review and meta-analysis included 35 studies with 13,854 participants. A Mallampati class above II was associated with 2.23 times the odds of being in the higher-AHI group. However, the studies used different cutoffs, with "high" AHI beginning anywhere from 5 to 15 events per hour 6.
This is an association, not a diagnostic rule. An odds ratio compares groups. It does not tell an individual whether they have OSA, how severe it is, or whether a sleep study can be skipped.
A prospective study of 137 adults referred for possible OSA found that each one-class increase was associated with 2.5 times the odds of OSA and an average increase of 5.2 AHI events per hour after adjustment for other measured variables 7. The confidence intervals were wide, and all participants had already been referred to a sleep center.
A larger sleep-clinic study reached a less useful clinical result. Among 953 adults, class IV had 40% sensitivity and 67% specificity for severe OSA. In plain language, most people with severe OSA were not identified by class IV, and class IV also occurred in people who did not have severe OSA. Class I was likewise poor at identifying people without OSA 3.
An earlier systematic review of eight studies found that separating suspected patients into classes I to II and III to IV changed their estimated probability of OSA only slightly and did not reliably rule the condition in or out 8.
These findings are not truly contradictory. A feature can be associated with OSA across a population yet still perform poorly as a test for one person. Differences in referral patterns, class definitions, examination technique, AHI thresholds, and participant characteristics also produce heterogeneity between studies.
Sensitivity and specificity in practical terms
Sensitivity asks how many people who have the target condition are identified by a test finding. Low sensitivity means the finding misses many cases. Because a high Mallampati class has limited sensitivity for OSA, having class I or II cannot rule OSA out.
Specificity asks how many people without the target condition do not have the test finding. Even a reasonably specific finding may be too weak to establish a diagnosis when used alone. Its usefulness also depends on how likely the condition was before the exam.
This is why a clinician may record the score as one clue without using it as the deciding factor.
What should guide an OSA evaluation instead
Ask a healthcare professional about OSA if you have symptoms such as frequent loud snoring, witnessed pauses in breathing, gasping during sleep, or persistent daytime sleepiness or fatigue 9. A partner's observations can be especially useful because many people do not know that their breathing changes during sleep.
A complete evaluation places the Mallampati finding beside:
- the pattern and duration of symptoms
- witnessed breathing pauses, snorting, or gasping
- daytime sleepiness and its effect on safety and daily function
- body mass index and neck size
- tonsils, tongue, jaw, palate, nasal airway, and other anatomy
- blood pressure, relevant health conditions, and medication use
- a validated multivariable screening tool when appropriate
These factors help form a pretest assessment, but no single symptom, measurement, or questionnaire establishes the diagnosis 101.
A questionnaire can help organize risk, but it still cannot diagnose OSA. The American Academy of Sleep Medicine recommends that clinical tools, questionnaires, and prediction algorithms not be used as substitutes for sleep testing in adults 10.
Sleep studies establish the diagnosis
In adults, polysomnography in a sleep laboratory is the standard diagnostic test. A technically adequate home sleep apnea test may be appropriate for an uncomplicated adult whose symptoms and evaluation indicate increased risk of moderate to severe OSA. The test should be selected and interpreted within a clinical evaluation, not chosen from a Mallampati class alone 10.
If a single home test is negative, inconclusive, or technically inadequate but OSA remains a concern, the guideline recommends polysomnography. It also recommends polysomnography rather than a home test for people with certain complicating conditions, including significant cardiorespiratory disease, possible sleep-related hypoventilation, neuromuscular respiratory weakness, chronic opioid use, a history of stroke, or severe insomnia 10.
Children need a pediatric pathway
Adult screening and home-testing guidance should not be applied to a child by default. Snoring, mouth breathing, breathing pauses, restless sleep, attention or behavior changes, and other concerns should be discussed with the child's pediatric clinician. The American Academy of Pediatrics guideline uses a separate diagnostic and management pathway and recommends polysomnography or specialist evaluation when a child snores regularly and has symptoms or signs of OSA 11.
The score does not assess central sleep apnea
Mallampati class concerns visible upper-airway anatomy, so it does not assess central sleep apnea. Central events involve problems with the brain's control of breathing rather than an upper-airway blockage, and obstructive and central events can occur in the same person 12. A sleep study and clinical evaluation distinguish these patterns.
What the score means before anesthesia
The anesthesia question is separate from the sleep apnea question. Direct laryngoscopy is the procedure used to view the larynx, while tracheal intubation is the placement of a breathing tube. Difficulty with laryngoscopy, tube placement, face-mask ventilation, or a supraglottic airway are related but distinct problems 5.
A pooled 2024 analysis found that the modified Mallampati test had 39% sensitivity and 86% specificity for difficult tracheal intubation in adults without obvious airway abnormalities. Results varied substantially among studies, and no single preoperative test was clearly superior 13. A 39% sensitivity means the Mallampati test alone missed most difficult intubations in the pooled data.
The American Society of Anesthesiologists guideline places Mallampati or uvula visibility within a broader assessment that includes airway history, mouth opening, teeth and jaw features, neck characteristics, head and neck movement, and other relevant findings 5.
Tell the anesthesia team before a procedure if you have:
- a previous difficult intubation, difficult ventilation, or airway emergency
- an airway alert letter, card, bracelet, or relevant anesthesia record
- diagnosed or suspected OSA
- new swelling, infection, injury, surgery, or limited movement involving the mouth, jaw, neck, or throat
The team can then assess the current airway and make a plan. A high class does not by itself dictate a particular device or technique, and a low class does not remove the need for an airway assessment.
Should treatment try to lower the score?
No treatment should be chosen simply to lower a Mallampati class. The score is a description, not a disease target. It does not show which tissue collapses during sleep, how often breathing is disturbed, or which therapy is most likely to work.
A visible class can change if oral or throat anatomy changes or if the exam is performed differently. That does not make "improving the score" a valid treatment goal. OSA treatment should follow a confirmed diagnosis and consider symptoms, sleep-study results, anatomy, other health conditions, preferences, and response to treatment.
When safety takes priority
If sleepiness makes it hard to keep your eyes open or maintain attention, do not start or continue driving. Arrange another ride or stop somewhere safe. Coffee alone cannot reliably compensate for serious sleepiness 14.
A Mallampati score is not an emergency finding. Sudden or severe difficulty breathing while awake, inability to speak normally because of breathlessness, blue or gray lips, confusion, collapse, or rapidly worsening swelling of the face, tongue, or throat requires emergency help 15.
The practical takeaway
The most accurate way to read a Mallampati score is as one line in a larger clinical picture. Classes III and IV can support concern about a crowded oral airway, but they do not establish OSA, measure its severity, or guarantee a difficult intubation. Classes I and II provide no safe reassurance that either problem is absent.
For possible OSA, let symptoms and a comprehensive evaluation guide clinician-directed sleep testing. For anesthesia, share previous airway and sleep-apnea information so the team can perform a current, complete airway assessment.
