The Owlet oxygen monitor is reasonably accurate for a consumer device, but it falls short of hospital-grade equipment in one critical area: catching low oxygen events as they happen. In clinical testing, the Owlet’s oxygen readings landed within about 2 to 2.5 percentage points of hospital monitors on average, which meets FDA standards. But when oxygen levels dropped dangerously low, the device missed a significant number of those episodes, raising questions about how much parents should rely on it for safety.
How Close Are the Readings to Hospital Monitors?
The most important accuracy metric for any pulse oximeter is something called root mean square error, which captures how far the device’s readings drift from a hospital-grade reference. In Owlet’s FDA submission for the Dream Sock, the at-home study showed an average error of 2.16% for oxygen saturation and 3.53 beats per minute for pulse rate. An earlier independent study of the Smart Sock found a similar error of 2.49%. To put that in context: if your baby’s true oxygen level is 97%, the Owlet might read anywhere from about 95% to 99% at any given moment.
The FDA requires pulse oximeters to stay below a 3% error threshold for oxygen readings and below 5 beats per minute for pulse rate. Owlet’s Dream Sock met both of those benchmarks in testing, which is why it received FDA clearance as a Class II medical device in November 2023. So for routine monitoring of a healthy baby, the readings are generally in the right ballpark.
The Real Problem: Detecting Low Oxygen
Accuracy of steady-state readings and accuracy of catching dangerous drops are two very different things. This is where the Owlet’s track record gets more complicated, and it’s the part that matters most to parents buying it for peace of mind.
A study published in Archives of Disease in Childhood tested the Owlet against a hospital pulse oximeter and found that when oxygen saturation fell below 80% for three or more seconds, the device’s sensitivity was just 14% using its smoothed (app-displayed) data. That means it caught only about 1 in 7 of those dangerous desaturation events. When researchers looked at the raw, unsmoothed data from the sensor itself, sensitivity jumped to 74%, but that raw data isn’t what parents see on their phones. The specificity was excellent at 100% for smoothed data, meaning false alarms for severe drops were essentially nonexistent. But a monitor that rarely gives false alarms because it rarely alarms at all isn’t very reassuring.
A separate study covered by JAMA found somewhat better results: the Owlet Smart Sock 2 showed 88.8% sensitivity and 85.7% specificity for detecting low oxygen. However, even in that study, 5 out of 12 patients who experienced hypoxemia had at least one episode where the Owlet displayed a normal reading while the hospital monitor showed dangerously low oxygen. The device did eventually flag each of those patients, but it didn’t catch every individual episode in real time.
Why the Gap Between Reading Accuracy and Event Detection
The disconnect comes down to how the Owlet processes its data before showing it to you. Consumer pulse oximeters smooth their readings to reduce the constant stream of minor fluctuations that would trigger nonstop false alarms. This smoothing averages out brief dips, which makes the display more stable but can mask short drops in oxygen. Hospital monitors in neonatal units are calibrated differently, with clinical staff interpreting raw waveforms and setting alarm thresholds specific to each patient.
The Dream Sock originally displayed only a 10-minute average oxygen saturation, which smoothed the data even further. After receiving FDA clearance in late 2023, Owlet began offering live oxygen readings in the U.S. version of the app. Canadian users still see only the 10-minute average. Whether this change to live readings improves real-world event detection hasn’t been independently tested yet.
Smart Sock vs. Dream Sock: What Changed
Owlet has sold several versions of its monitor, and the differences matter for accuracy. The Smart Sock 3 displayed live oxygen and pulse rate readings and sent red alerts when values fell outside preset ranges. The FDA issued a warning letter about this product in 2021, arguing it was functioning as an uncleared medical device. Owlet pulled the Smart Sock 3 from the market.
The Dream Sock launched as a “wellness device” without the red alerts, sidestepping the regulatory issue. Then in November 2023, Owlet received formal FDA clearance as a Class II over-the-counter infant pulse oximeter, allowing the company to reintroduce health notifications. This makes the Dream Sock the first consumer infant pulse oximeter with FDA clearance, which means it has met a defined accuracy standard. But FDA clearance doesn’t mean the device performs identically to clinical monitors. It means it met the minimum threshold for its device category.
Factors That Affect Accuracy
Several practical issues can push the Owlet’s accuracy lower than what clinical studies report. Fit is the most common one. The sensor needs consistent contact with the skin on your baby’s foot, and if the sock shifts, bunches, or is the wrong size, readings become unreliable or drop out entirely. Many parents report frequent disconnections or yellow notifications (indicating the device lost a good signal) when the sock isn’t positioned correctly.
Movement is another factor. Babies kick, squirm, and curl their toes, all of which can disrupt the optical sensor. Owlet’s FDA testing showed the device met accuracy thresholds during both motion and non-motion conditions, but controlled motion simulation in a lab is different from a restless infant at 3 a.m.
Skin tone is a broader concern across all pulse oximeters, not just Owlet. The FDA has acknowledged that current evidence shows accuracy differences between individuals with lighter and darker skin pigmentation. Pulse oximeters work by shining light through tissue, and melanin absorbs some of that light, which can affect readings. The FDA has proposed updated testing recommendations to address this, and has partnered with academic institutions on studies in pediatric patients with varying skin tones. For now, parents of darker-skinned infants should be aware that readings may skew slightly higher than actual oxygen levels, a pattern well documented in adult hospital oximeters.
What This Means in Practice
The Owlet is accurate enough to give you a general picture of your baby’s oxygen levels during sleep. If your baby is healthy and breathing normally, the readings will reflect that. Where it falls short is in the scenario many parents buy it for: catching a sudden, dangerous drop in oxygen that might indicate a breathing emergency. Independent testing suggests the device will miss some of those events, particularly brief desaturations, because of how it smooths and averages data.
It’s worth noting what the Owlet is not designed to do. It is not a replacement for a hospital apnea monitor, and it is not intended to detect or prevent sudden infant death syndrome (SIDS). No consumer monitor has been shown to reduce SIDS risk. The American Academy of Pediatrics does not recommend home pulse oximetry for healthy infants, partly because of the limitations described above and partly because false alarms (or worse, false reassurance) can cause more harm than good.
For parents of infants with known respiratory or cardiac conditions, medical-grade home monitors prescribed by a physician offer tighter accuracy and clinically validated alarm thresholds. The Owlet occupies a middle ground: better than no data, but not reliable enough to be your only safety net.