The Mallampati Score: What a Clinician Sees When You Open Your Mouth
Open wide and stick your tongue out, and a clinician grades how much of your throat is hidden. It is the cheapest airway assessment there is, it takes five seconds, and the study most often cited for it is weaker on the point that matters than its own conclusion suggests.
It is the fastest airway assessment in medicine. You open your mouth, put your tongue out, and someone grades how much of the back of your throat they can still see. Four classes. No equipment. Five seconds.
It is genuinely useful, and it is routinely oversold — including, as it happens, by the conclusion of the study most often quoted in its favour.
What does the Mallampati score measure?
Crowding. Specifically, how much space your tongue leaves in your mouth, which is a proxy for how much space there is in the airway behind it.
The modified Mallampati classification runs in four classes, from a clear view of the soft palate, tonsillar pillars and uvula at Class I, through to Class IV where only the hard palate is visible and the tongue obscures everything behind it.
The logic maps directly onto this site's organising question. A high Mallampati class says the obstruction risk sits at the tongue base and oropharynx, not at the nose. That is a different level of the airway, and it predicts which interventions can physically do anything. A nasal strip opens the nostril; it does nothing whatsoever about a crowded throat. Nose, mouth or throat snoring.
The score originated in anaesthesia, as a way of predicting a difficult intubation. Its use in sleep medicine is borrowed, and some of the overconfidence comes from that transfer.
How is Friedman staging different?
Friedman tongue position staging is the related system you will see alongside it. It grades the same crowding but combines the palate view with tonsil size and body mass index, which makes it more specifically a sleep-apnea instrument rather than an airway-access one.
The practical difference: Mallampati is about the tongue filling the mouth, Friedman adds what the tonsils are doing. In someone with large tonsils those are the obstruction, and the tonsils are what should be addressed. Enlarged tonsils in adults.
What did the study actually find?
This is where the usual write-up stops reading too early.
Nair and colleagues, in Cranio in 2025, gave the STOP-BANG questionnaire to 145 adults. Of the 95 who screened as high risk, 42 went on to have a full polysomnogram, and 24 of those were confirmed to have obstructive sleep apnea.
Two separate comparisons came out of that, and they did not agree.
Against questionnaire risk, the scores performed well. 86% of the low-risk group had only Class I Mallampati. In the high-risk group, Class III appeared in 37.9% and Class IV in 17.9%, at p < .001. For Friedman staging, every low-risk participant was Stage 1, while Stages 3 and 4 appeared in 27.4% and 1.1% of the high-risk group, also p < .001.
Against confirmed apnea on polysomnography, they did not. In the 24 participants with PSG-proven OSA, Class III and IV appeared in 50% and 20.8% — proportions that look impressive until you read the p-value the authors report alongside them: p = .831. For Friedman staging in the same group, p = .092.
The paper's own conclusion is that "either MMP or FS scores along with the STOP-BANG questionnaire can be used to diagnose OSA" in limited settings. That is a more confident sentence than the second comparison supports.
What it most likely demonstrates: Mallampati class tracks the things STOP-BANG asks about, and this sample was too small — 24 confirmed cases — to establish an independent association with apnea itself. Absence of significance in 24 people is weak evidence of absence. It is also not the positive finding it is usually cited as.
Evidence tier 2. The anatomical reasoning is sound and a high class genuinely tells you where to look. Its value as a standalone predictor of confirmed apnea is not established by this study, whatever its abstract says.
Can you check your own?
Roughly, and with a caveat.
Open your mouth wide in front of a mirror, tongue out, without saying "ahh" — phonating lifts the palate and flatters the result, which is one of the ways the score drifts between examiners.
- Class I — you can see the soft palate, the uvula and the pillars either side
- Class II — soft palate and uvula, pillars hidden
- Class III — soft palate and only the base of the uvula
- Class IV — hard palate only
The caveat: inter-observer agreement on this is mediocre even between clinicians, and it varies with head position, tongue protrusion and whether you phonate. Treat your own result as a hint about where to investigate, never as a score.
A Class III or IV finding is a reason to take snoring seriously and to mention the crowding when you are assessed. It is not a diagnosis, and a Class I does not clear you — plenty of people with significant apnea have roomy-looking mouths, because their narrowing is somewhere the torch does not reach.
What should you do with the answer?
If it is high, say so at the appointment. It directs attention to the tongue base and oropharynx, and it is relevant to whether an oral appliance or tongue-directed treatment is likely to help. What to tell the sleep doctor.
Do not substitute it for a sleep study. No mouth examination counts breathing events or measures oxygen. The crowding tells you about anatomy; the study tells you what that anatomy is doing at 3am. Is it worth seeing a doctor about snoring.
And do not buy from it. A high Mallampati score tells you the narrowing is behind your tongue, which is the single clearest reason a nasal product will not help you — a result worth having precisely because it saves money rather than directing it.
No consumer product treats obstructive sleep apnea. If you have witnessed pauses, gasping arousals or unrefreshing sleep, the examination is a step on the way to assessment rather than a substitute for it. Snoring versus sleep apnea.