CPAP Mask Leaks & Mouth Leak: Causes, the Data Threshold, and Fixes

Updated 2026-06-21 10 min read

Mask and mouth leaks ruin comfort and corrupt your AHI. Learn the leak thresholds, why a low AHI can be a lie, and how to fix common leaks.

A small amount of leak is normal — your mask is designed to leak. The problem starts when air escapes where it shouldn't: around a poorly sealed cushion or out of your mouth while you sleep. Beyond ruining comfort and drying you out, a large leak can quietly corrupt the very numbers you rely on, making a "good" AHI look better than your night actually was. This guide explains the leak you should ignore, the leak that matters, the exact threshold, and how to fix the most common causes.

This article is part of our Troubleshooting & Optimizing CPAP pillar. It's data education, not medical advice — discuss any therapy changes with your sleep clinician.

Intentional vent leak vs unintentional leak

Every CPAP mask has tiny holes — the diffuser vent — engineered to bleed off your exhaled carbon dioxide so you don't rebreathe it. That airflow is the intentional leak (sometimes called the vent leak). It scales with pressure: the higher your pressure, the more air the vent releases by design. This is expected and healthy.

Unintentional leak is everything else — air escaping from a spot the mask was never meant to let go:

  • A cushion that has shifted, folded, or worn out
  • A strap that's too loose (or, paradoxically, too tight and buckling the seal)
  • A mask the wrong size or shape for your face
  • Your mouth falling open while you wear a nasal mask or nasal pillows

Here's the catch that trips people up: manufacturers report leak differently.

  • ResMed reports excess leak — it subtracts the expected vent leak and shows only the unintentional portion. A ResMed number close to zero means a great seal.
  • Philips reports total leak — vent leak plus any unintentional leak, lumped together. Philips numbers look "higher" for the same real-world seal because they include the by-design venting.

Because the baselines differ, you cannot compare a raw ResMed leak figure to a raw Philips one. This is one of the most common reasons people panic over a "high" number that's actually fine — or shrug off a number that isn't. (For a deeper dive on what counts as acceptable, see CPAP Leak Rate: What's Acceptable and How to Fix High Leaks.)

The threshold that matters (24 L/min on ResMed)

For ResMed machines, the widely used red line is about 24 liters per minute (L/min) of excess leak at the 95th percentile. The 95th-percentile figure means: for 95% of your night, your excess leak stayed at or below this value. It's a trend statistic, not a momentary spike, which makes it far more meaningful than the brief blips that happen every time you roll over.

Metric What it tells you
Median leak Your typical, steady-state seal
95th-percentile leak The level you stayed under for most of the night
ResMed "Large Leak" flag 95th-percentile excess leak above ~24 L/min

If your ResMed report crosses ~24 L/min at the 95th percentile, the machine flags the night as having a large leak. On Philips devices the displayed total-leak number is higher by design, so the same conceptual threshold sits at a different raw value — yet another reason not to compare the two brands' numbers directly. To understand how the 95th-percentile statistic works across metrics, see CPAP 95th-Percentile vs Median Pressure: What the Numbers Mean.

The reason this single number deserves your attention is what it does to the rest of your data.

Why large leak falsely under-reports AHI

Your CPAP estimates apneas and hypopneas by watching your flow rate — the airflow signal in and out of the mask. When a large leak is dumping air out of a gap or your mouth, the machine can no longer "see" your breathing cleanly. The result is counterintuitive but important:

A large leak can cause your machine to miss events and under-report your AHI. Your number can look better during a leaky night, not worse.

The machine essentially loses confidence in the flow signal during a heavy leak, and many devices won't score events it can't reliably detect. So the leak doesn't just degrade comfort — it corrupts the data, including subtler metrics like flow limitation, the early-warning sign of airway narrowing that lives beneath your AHI.

This is the trust problem at the heart of CPAP self-monitoring: a low AHI on a high-leak night is not reassurance — it's an asterisk. Remember, the treatment goal for adults on CPAP is generally a residual AHI below 5 events per hour; many clinicians aim lower (e.g., under 1–2) when comfortably achievable, though targets are individualized with your provider. None of those targets mean anything if the underlying flow data was compromised by leak. (More on interpreting the number itself in What Is a Good AHI on CPAP?)

This is exactly where automatic analysis earns its keep. SomniCharts flags large-leak nights automatically across brands — ResMed, Philips Respironics (including the encrypted DreamStation 2), and Löwenstein prisma — and tells you in plain language whether a spike was a mask-seal leak or a mouth leak, so you know whether a good AHI is real or just an artifact of air escaping.

Fixing seal leaks — refit, cushion, headgear

Most seal leaks come from fit, not faulty hardware. Work through these in order before assuming you need a new mask:

  1. Refit while lying down, with the pressure on. Your face changes shape when you lie back. Put the mask on, start the machine, then gently lift the cushion away from your face for a second and let it re-seat — this lets the air pressure inflate the cushion against your skin (the "float and settle" trick).
  2. Loosen, don't crank. Over-tightening is the #1 cause of leaks. Too-tight straps deform the cushion, opening gaps and digging in. Snug, not strangling.
  3. Inspect the cushion. Silicone cushions wear out, harden, and tear. Most need replacing every 1–3 months. A glossy, stiff, or visibly cracked cushion will leak no matter how you adjust it.
  4. Check headgear elasticity. Stretched-out straps can't hold tension. Headgear typically lasts 3–6 months before it needs replacing.
  5. Reconsider the mask type. Nasal pillows leak differently than full-face masks; facial hair, a high nasal bridge, or sleeping position can all dictate which style seals for you. See CPAP Mask Types and Fit: Nasal, Pillow, and Full-Face Compared.

A quick reference for the usual culprits:

Symptom Likely cause First fix
Leak from the top, air in eyes Mask too high / too tight at forehead Reposition lower, loosen top straps
Leak at cheeks Wrong size or worn cushion Resize or replace cushion
Leak worsens over the night Straps stretching, cushion sliding New headgear; float-and-settle refit
Leak only when on your side Pillow pressing the mask Try a CPAP pillow or different mask style

If you keep crossing the leak threshold despite a good fit, persistent leak can be a sign the mask simply doesn't match your face — a conversation worth having with your equipment provider (DME).

Mouth leak — chin straps and full-face options

Mouth leak is a different animal. Your seal can be perfect, yet air drives in through your nose and rushes out your open mouth — a one-way river of pressurized air that dries out your mouth and throat and disrupts therapy. It's most common with nasal masks and nasal pillows, and it's a frequent reason people wake up parched. (If dry mouth is your main complaint, start with CPAP Dry Mouth and Mouth Breathing: How to Stop It Tonight.)

Two main approaches:

  • A chin strap gently holds your jaw closed to keep your lips sealed. It's inexpensive and works well for people who are mostly nasal breathers but drift open in deep sleep. It does little if you're a committed mouth breather or are congested.
  • A full-face mask sidesteps the problem entirely by covering both nose and mouth, so an open mouth no longer means a leak. This is often the more reliable fix for chronic mouth breathers, people with nasal congestion, or anyone for whom a chin strap doesn't hold. (See CPAP Nasal Congestion and Stuffy Nose: Causes and Fixes if a stuffy nose is forcing your mouth open.)

A practical note on data: mouth leak often shows up as a leak that climbs later in the night, as you sink into deeper, more relaxed sleep — distinct from a seal leak that's there from the start or appears with position changes. Knowing which pattern you have points you to the right fix, which brings us to the most useful way to read leak.

Reading leak as a timeline

A single overnight leak number — even the 95th-percentile one — is a summary that hides the story. The real diagnostic power is in the leak graph over time, the minute-by-minute trace across your night. Single-night numbers are noise; trends over weeks are signal.

Reading leak as a timeline lets you tell the causes apart:

  • Seal leak often appears as sudden spikes tied to movement or position changes, or a leak that's elevated from the moment you put the mask on.
  • Mouth leak tends to ramp up gradually, frequently in the second half of the night as your jaw relaxes.
  • Mask-off / removal events show as a flat-topped, very high plateau when the mask comes off entirely — covered in Mask Off & Large-Leak Events: Why Your CPAP Report Shows Gaps.

Cross-referencing the leak timeline against your AHI timeline is what exposes the "good AHI is a lie" trap: if your events drop to near-zero exactly when your leak spikes, that low count is an artifact, not a triumph.

This pattern-matching is tedious to do by hand night after night — which is the everyday job SomniCharts automates. It imports your ResMed, Philips (including DreamStation 2), or Löwenstein prisma data, plots leak as a readable timeline alongside your events and pressure, and labels whether each spike looks like seal or mouth leak — so you walk into your next appointment knowing whether your numbers can be trusted. For the bigger picture on interpreting every metric on your report, see How to Read Your CPAP Data: A Plain-English Guide to AHI, Leak Rate & Pressure.

Frequently asked questions

What is a good CPAP leak rate? On ResMed (excess leak), staying under about 24 L/min at the 95th percentile keeps you out of large-leak territory; close to zero is ideal. Philips reports total leak, so its numbers run higher by design and can't be compared directly to ResMed's.

Why does my CPAP mask leak even when it's tight? Over-tightening is a leading cause of leaks — it deforms the cushion and opens gaps. Try loosening, refitting while lying down with the pressure on, and replacing a worn cushion before tightening further.

Can a leak make my AHI look better than it really is? Yes. A large leak degrades the flow signal the machine uses to detect events, so it can under-report your AHI. A low AHI on a high-leak night should be treated with caution, not celebrated.

Will a chin strap fix mouth leak? Often, if you're mostly a nasal breather who drifts open in deep sleep. Dedicated mouth breathers, or anyone with nasal congestion, usually do better with a full-face mask.

Frequently asked questions

Can a mask leak make my AHI look better?

Yes. A large leak can stop the machine from scoring events, falsely lowering your AHI — which is why leak and AHI should be read together.

Turn your CPAP data into answers

SomniCharts imports your ResMed, Philips Respironics, or Löwenstein data and automatically explains your AHI, leaks, and pressure — no spreadsheets, no OSCAR setup.

7-day free trial · cancel anytime

References

  1. CPAP Mask Leaks: Causes & How to Fix Them — ResMed
  2. OSCAR leaks — Apnea Board Wiki

This article is for general education and is not medical advice. Always consult a qualified clinician about your therapy. See our Medical & Clinical Disclaimer.

← Back to all guides