The Mask Makes the Therapy: A 2026 Guide to Choosing the Right CPAP Mask — and Reading Your Leak Data
The machine gets the attention, but the mask is where CPAP succeeds or fails. Dr. Vishal Saini reviews how CPAP masks have quietly transformed over the last few years, which mask suits which patient, what the research actually says about mask choice, why fit and comfort aren't luxuries but determinants of whether the therapy works at all — and the question he's asked constantly: when should you worry about your leak number, and when can you safely ignore it?
By Dr. Vishal Saini, M.D., FAASM — Mid-West Center for Sleep Disorders
Patients spend weeks researching which CPAP machine to get. Then they strap on whatever mask the setup technician happened to grab, discover it hisses, digs into the bridge of the nose, or blasts air into their partner's face, and conclude that "CPAP doesn't work for me." I want to push back on that, hard, because in my experience the single most common reason CPAP fails isn't the machine, the pressure, or the diagnosis. It's the mask. Get the interface right and a reluctant patient becomes a nightly user; get it wrong and the best machine in the world sits in a closet.
The good news is that this is one of the fastest-moving corners of sleep medicine. The masks available in 2026 are dramatically better — lighter, quieter, less intrusive — than what we were fitting even five years ago. So let's walk through the landscape: the types, the recent advances, who each one suits, what the evidence shows, and how to actually read the leak data your machine sends you every morning.
The Three Families of Mask
Nearly every CPAP mask belongs to one of three categories, and understanding them is the foundation of every good choice.
Nasal pillow masks seal at the nostrils with two small soft "pillows" that rest just inside the nose. They're the most minimal option — tiny, light, with almost nothing in your field of view — and they're excellent for people who feel claustrophobic, who read or watch TV before sleep, who wear glasses, or who have facial hair that defeats a larger cushion. Their limit: at very high pressures the jet of air at the nostrils can feel intense or dry, so they're usually best at low-to-moderate pressures.
Nasal masks cover the nose with a triangular cushion sealing around it. They're the long-standing workhorse and, as you'll see below, the type the research tends to favor. They distribute pressure comfortably across a wider range and suit most people who breathe through their nose during sleep.
Oronasal (full-face) masks cover both the nose and mouth. They exist for a specific reason: patients who are confirmed mouth-breathers, who have chronic nasal congestion or obstruction, or who simply can't keep their mouth closed on a nasal mask and leak badly as a result. They are a legitimate and sometimes necessary tool — but, importantly, they are not the default, and they shouldn't be the automatic first choice just because they "cover everything."
"Patients assume the full-face mask is the 'more complete' option because it covers more of your face. It isn't. For most people with sleep apnea, a nasal interface is the better starting point, and the data back that up. The full-face mask is a solution to a specific problem — mouth breathing or nasal blockage — not a one-size-fits-all upgrade." — Dr. Vishal Saini
What's New: How Masks Have Changed in the Last Few Years
If your mask is more than a couple of years old, you may not recognize what's on offer now. A few genuine advances stand out.
Top-of-head tube routing. A wave of designs now run the air tube up and over the top of the head rather than straight out the front (ResMed's N30i and P30i nasal and pillow masks, the F30i full-face, and similar designs from other makers). The hose is out of your way, so you can sleep on your side or stomach and roll over without the tube tugging the seal loose. For active sleepers this has been a quiet revolution.
Minimal-contact and "cradle" cushions. Newer nasal masks (such as ResMed's AirFit N30 and Fisher & Paykel's Evora) tuck under the nose rather than over the bridge, eliminating the red marks and pressure sores at the top of the nose that drove so many people crazy — and getting the mask out of your sightline entirely.
Softer, smarter cushions. Memory-foam cushions (like the AirTouch line) mold to the face instead of pressing back against it, and self-conforming designs aim to reduce the fiddly manual fitting that used to be a nightly chore. Headgear has gotten lighter and more adjustable, and magnetic clips make the mask easier to take on and off — though a note of caution there: some magnetic headgear clips carry warnings for people with certain implanted medical devices (pacemakers, certain neurostimulators, metallic implants near the head), so flag any implant to your sleep team before using a magnetic mask.
The through-line in all of this is less mask, more comfort — and as I'll argue, comfort is not a frill. It's the mechanism by which the therapy works.
Because product models change constantly and I can't show you copyrighted product photos here, the best way to actually see the current options is to browse the manufacturers' own mask pages: ResMed masks and Fisher & Paykel masks both keep current catalogs with photos, sizing, and fit guides. Your durable medical equipment (DME) supplier can show you the specific models they stock.
Which Mask for Which Patient
Here's how I think it through in clinic.
Claustrophobic, or want the least mask possible? Start with nasal pillows. The minimal footprint is often the difference between tolerating therapy and abandoning it.
Nose-breather on low-to-moderate pressure, glasses or bedtime reading? A nasal pillow or an under-nose nasal cushion is usually ideal.
Nose-breather needing a reliable all-around seal, or on higher pressure? A standard nasal mask spreads the pressure comfortably and is the dependable default for most people.
Side or stomach sleeper, or someone who tosses and turns? Look specifically at a top-of-head tube ("...i"-style) design so the hose doesn't break your seal when you move.
Confirmed mouth-breather, chronic nasal congestion, or persistent mouth leak on a nasal mask? This is the real indication for an oronasal (full-face) mask — but before defaulting to one, it's worth asking whether the nasal problem itself can be treated (allergy/congestion management, treating a deviated septum) or whether a chin strap plus a nasal mask would solve the mouth-opening, because, as the data below show, staying nasal has real advantages.
Facial hair or an unusual facial structure? Nasal pillows often seal better than a cushion that has to bridge over a beard; otherwise a fitting session trying several styles is worth the time.
What the Research Actually Says About Mask Choice
This isn't just preference — there's a real and surprisingly consistent evidence base, and it points in one direction.
The American Thoracic Society thought mask selection important enough to convene a formal workshop on it. Its report concluded that the interface materially affects CPAP outcomes — adherence, residual apnea, leak, and the pressure required — and called for personalized mask selection with the patient (and bed partner) actively involved, rather than whatever happens to be on the shelf (Genta et al., Annals of the American Thoracic Society, 2020, PMID 33000960).
When you compare nasal against oronasal masks head-to-head, the nasal route wins on the measures that matter. A meta-analysis of 13 trials and more than 4,500 patients found that, compared with nasal masks, oronasal masks were associated with higher required CPAP pressure, higher residual apnea-hypopnea index (AHI), and lower adherence (Andrade et al., Chest, 2018, PMID 29273515). A more recent network meta-analysis of 29 studies and over 6,300 patients reached the same conclusion: nasal masks and nasal pillows were associated with lower residual AHI and lower required pressure than oronasal masks (Chen et al., Journal of Sleep Research, 2022, PMID 35821391). And in a controlled crossover trial where the same patients used the same pressure with each mask, residual breathing events were significantly higher on the oronasal than the nasal mask (Ebben et al., Sleep Medicine, 2014, PMID 24831252) — meaning the mask itself, not the patient, changed how well the therapy worked.
The practical message: a nasal interface is the right starting point for most patients, and a full-face mask is a considered choice for a real reason — not a reflex. If you're on a full-face mask and your residual AHI is stubbornly high or your pressures keep climbing, the mask is one of the first things I re-examine.
Why Fit and Comfort Are Not Luxuries
Here's the idea I most want to land: comfort isn't the opposite of effectiveness — comfort is effectiveness, because CPAP only works on the hours you actually wear it.
CPAP is a remarkably effective therapy, but its real-world benefit is capped by adherence, and adherence is driven heavily by the interface and its side effects — pressure marks, air in the eyes, noise, claustrophobia, a seal that fails when you move (Sawyer et al., Sleep Medicine Reviews, 2011, PMID 21652236). A mask that's 5% more comfortable but that you wear two extra hours a night delivers far more treated sleep than a "better-sealing" mask you rip off at 3 a.m. The math of this is lopsided, and it's why I'd rather spend a clinic visit on mask fit than almost anything else.
There's a second, subtler reason fit matters: a poorly fitting mask doesn't just annoy you — it leaks, and leaks degrade the therapy itself, both by letting pressure escape and by throwing off the machine's ability to detect and treat your events. Which brings us to the question I'm asked more than any other.
Reading Your Leak Data: When to Worry, and When to Let It Go
Modern machines and their apps (myAir, SleepMapper, and the clinician-facing dashboards) report a nightly leak number, and patients fixate on it — often more than they should. Let me demystify it.
First, the crucial distinction: there are two kinds of leak. Every mask is designed to leak — it has vent holes that continuously flush out the carbon dioxide you exhale. That's intentional (vented) leak, and it's supposed to be there. What your machine is really trying to flag is unintentional leak — air escaping where it shouldn't, around a loose seal or an open mouth. Many devices report total leak (intentional plus unintentional lumped together), which is why the raw number can look alarming when nothing is actually wrong.
A few principles I give patients:
Total leak naturally rises with pressure — don't panic at a bigger number. The higher your pressure, the more air flows out the intentional vents. A patient on 14 cm of water will show a higher total-leak number than one on 7, with both masks sealing perfectly. The absolute number means little without context.
Judge leak by its company, not in isolation. The question is never "is my leak number high?" in a vacuum. It's "is my leak high and is my residual AHI rising, and do I have symptoms?" If your residual AHI is well controlled, you wake rested, and you have no dry eyes, noise, or skin marks, then a leak figure that sits in your machine's acceptable ("in range" / green) zone is fine to leave alone — even if it's not zero. Chasing a perfect leak number you don't need is a recipe for overtightening the mask and creating new problems.
Worry when leak is high and one of these is true: your residual AHI climbs with it (the leak is defeating the therapy); you're waking with dry, irritated eyes or a dry mouth, or air is waking you or your partner; the leak consistently crosses your device's "large leak"/red threshold night after night (for many machines that unintentional-leak alarm sits around the mid-20s of liters per minute); or the number suddenly worsens from your own baseline, which usually signals a worn cushion, a cracked frame, or a changed fit. A persistent, symptomatic, AHI-wrecking leak is a real problem and a reason to call your sleep team.
Mouth leak is its own category — and often a mask-choice signal. If your leak spikes are coming from your mouth falling open during sleep (a mechanism recent research has tied directly to the jaw dropping open under the effort of breathing; Martinot et al., Annals of the American Thoracic Society, 2026, PMID 42608809), the fix isn't a tighter nasal mask — it's addressing the mouth, with a chin strap or by moving to a full-face mask.
And some leak is simply life. Brief spikes when you adjust the mask, shift position, or take it off in the morning are normal and show up in the data as transient blips. A single bad night after a cold, or one restless night, isn't a trend. Look at patterns over a week or two, not at one alarming morning readout.
"The leak number is a tool, not a verdict. I've watched patients lie awake cranking their headgear tighter to chase a lower figure, give themselves pressure sores, and sleep worse — when their therapy was working fine the whole time. Read the leak in context: alongside your residual AHI and how you actually feel. Those three together tell the real story." — Dr. Vishal Saini
The Bottom Line
The mask is not an afterthought to CPAP — it is the therapy's point of contact with you, and the evidence is clear that the choice matters: nasal masks and nasal pillows generally outperform full-face masks on residual apnea, required pressure, and adherence, so a nasal interface is the right default and a full-face mask is a deliberate choice for mouth breathing or nasal obstruction. The last few years have made the whole category lighter, quieter, and far less intrusive, which matters because comfort is precisely what converts a prescription into nightly use — and CPAP only treats the hours you wear it. As for your leak data: distinguish the intentional vent from true leak, remember the total number rises with pressure, and judge it alongside your residual AHI and how you feel. Worry when leak is high and your numbers or symptoms say so; otherwise, don't let a nonzero figure rob you of sleep. And if your mask is fighting you, don't conclude that CPAP has failed — come in and let's find the one that fits.
Dr. Vishal Saini, M.D., FAASM is the Research & Medical Director at Mid-West Center for Sleep Disorders, which operates its own sleep laboratory and performs attended CPAP titrations and mask fittings for patients with obstructive, central, and complex sleep apnea in Lansing, Traverse City, and Eaton Rapids.
Struggling with your CPAP mask, or staring at leak numbers you don't know how to read? Let's fit the right mask and make the therapy something you'll actually wear.
Book a Consultation → | (517) 887-6733
References: Genta PR, Kaminska M, Edwards BA, et al. The Importance of Mask Selection on Continuous Positive Airway Pressure Outcomes for Obstructive Sleep Apnea. An Official American Thoracic Society Workshop Report. Ann Am Thorac Soc 2020;17(10):1177–1185 (PMID 33000960). Andrade RGS, Viana FM, Nascimento JA, et al. Nasal vs Oronasal CPAP for OSA Treatment: A Meta-Analysis. Chest 2018;153(3):665–674 (PMID 29273515). Chen LY, Chen YH, Hu SW, et al. In search of a better CPAP interface: A network meta-analysis comparing nasal masks, nasal pillows and oronasal masks. J Sleep Res 2022;31(6):e13686 (PMID 35821391). Ebben MR, Narizhnaya M, Segal AZ, Barone D, Krieger AC. A randomised controlled trial on the effect of mask choice on residual respiratory events with continuous positive airway pressure treatment. Sleep Med 2014;15(6):619–624 (PMID 24831252). Sawyer AM, Gooneratne NS, Marcus CL, et al. A systematic review of CPAP adherence across age groups. Sleep Med Rev 2011;15(6):343–356 (PMID 21652236). Patil SP, Ayappa IA, Caples SM, et al. Treatment of Adult Obstructive Sleep Apnea with Positive Airway Pressure: An American Academy of Sleep Medicine Clinical Practice Guideline. J Clin Sleep Med 2019;15(2):335–343 (PMID 30736887). Martinot JB, Le-Dong NN, Clause D, Malhotra A, Pépin JL. Respiratory Effort-Driven Mandibular Opening as a Mechanism of Air Leak During CPAP Therapy. Ann Am Thorac Soc 2026 (PMID 42608809). Product examples are illustrative, not endorsements; see manufacturer mask pages (ResMed, Fisher & Paykel) for current models and fitting guides.