Skin temperature on wearables: why it is a deviation, not a thermometer reading
Your ring does not take your temperature. It tracks how your nightly skin temperature differs from your own normal. That is less than a thermometer and, for some uses, more.
Oura, WHOOP and several other devices now report a nightly temperature figure. It is usually shown as something like "+0.4 °C", not "36.9 °C". That is deliberate. These sensors measure the skin of a finger or wrist, which can swing by several degrees with room temperature, bedding and blood flow. What they can do well is compare tonight with your own baseline under similar conditions.
What the companies say they measure
Oura's help centre says it "measures average body temperature from the skin," during sleep "while your body is in its most stable state," and displays changes "as fluctuations (+/-) from your personal baseline." Oura calculates that baseline over the first couple of weeks of wear, and a new ring may take time to re-establish it. WHOOP lists skin temperature among the inputs to Recovery. Neither presents the figure as a core body temperature, and neither is a medical thermometer.
Why the deviation is useful
Core body temperature follows a daily rhythm and rises with fever. Skin temperature is noisier, but measured at the same point in the sleep cycle every night, the night-to-night difference becomes a sensitive marker of change, even though the absolute value is not a body temperature.
What the research found
| Study | Year, journal | Who and how many | Design | Main result |
|---|---|---|---|---|
| Smarr et al. | 2020, Scientific Reports | First 50 people in the TemPredict study reporting COVID-19 | Continuous finger temperature from Oura | Illness-related rises in skin temperature were visible and correlated with self-reported fever; some rises preceded symptom recognition (one author consulted for and one was employed by Oura) |
| Mason et al. | 2022, Scientific Reports | 63,153 participants; 73 PCR-confirmed cases for training | Machine learning on Oura data | Detected COVID-19 an average of 2.75 days before testing; sensitivity 82%, specificity 63%; adding continuous temperature raised the AUC by 4.9% |
| Maijala et al. | 2019, BMC Women's Health | 22 women, about 115 days each | Nightly finger skin temperature from Oura vs oral temperature and ovulation tests | Skin temperature 0.30 °C higher in the luteal phase (oral: 0.23 °C); menstruation detected with 71.9 to 86.5% sensitivity in plus or minus 2 to 4 day windows |
These studies show two things. Illness, especially with fever, shows up as a rise above baseline that a ring can see. And the small, regular shift across a menstrual cycle, about a third of a degree, is also visible in nightly skin temperature, which is why several devices offer cycle-related temperature features. Both are measured as deviations, and both depend on wearing the device consistently.
What else moves it
- Bedroom temperature and bedding. A hot room or heavy duvet raises skin temperature. A night in a hotel can shift it.
- Infection. The TemPredict studies above found illness-related rises in finger temperature.
- Evening habits. Alcohol, late meals and late exercise are commonly blamed. The controlled evidence on their effect on wearable skin temperature is thin, so log them and compare your own nights.
- Travel and time zones, because the measurement window shifts with your sleep.
- Fit. A loose ring or strap reads less reliably.
How the studies were built
The TemPredict study recruited more than 63,000 people early in the pandemic, asked them to complete daily symptom surveys and paired those with continuous ring data. Only 73 of the 704 people who reported possible COVID-19 had both PCR confirmation and high-quality data, and those were used to train the model. In a further 10 people confirmed by antibody testing, the model had 90 percent sensitivity and 80 percent specificity. The researchers also found substantial differences in accuracy by age and sex. The 2020 paper declares that one author had consulted for Oura and another was an Oura employee. In the menstrual cycle study, an algorithm using skin temperature detected ovulation with 83.3 percent sensitivity within a window from three days before to two days after the verified date.
Ring versus thermometer
The menstrual cycle study is one of the few that compared ring skin temperature directly with a conventional measurement. Daily finger skin temperature correlated with oral temperature taken on waking at r = 0.563, and the size of the shift between cycle phases correlated at r = 0.589. The ring detected the shift slightly more strongly than the thermometer (0.30 against 0.23 °C). Predicted menstruation dates were off by an average of 0.4 days (SD 1.8). The ring is not a substitute for a thermometer when you need an absolute temperature, but for a nightly, effortless trend it compares well. Oura notes that a new or replacement ring may take time to re-establish your baseline, so expect some noise after switching.
How to read a change
A single night at +0.3 °C means little. A rise of a similar size that persists, especially with a higher resting heart rate, lower HRV or a higher respiratory rate, is a more meaningful signal that something is going on, often an infection. A persistent rise with symptoms is a reason to use a real thermometer and, if needed, see a clinician. A wearable deviation is not a fever reading and cannot rule one out.
Oura's Readiness Score includes body temperature as a contributor, so a temperature rise will usually lower Readiness too. See Oura Readiness.
Logging it
Peptide AI reads nightly temperature deviation where the device provides it through wearable integrations and shows it with HRV, resting heart rate and respiratory rate. The dose log on the same axis helps tell a reaction to a new injection or vaccine from a cold coming on.
Questions people ask
Is wearable skin temperature my body temperature?
No. Devices such as Oura measure finger skin temperature during sleep and show it as a deviation from your personal baseline, not as a core body temperature.
What does a positive temperature deviation mean?
Your skin was warmer than your usual night. Common causes are a warm room, alcohol, a late meal, illness or the luteal phase of the menstrual cycle.
Can a ring detect a fever?
It can detect rises consistent with fever. In a 2022 study of 63,153 Oura users, a model including temperature detected COVID-19 an average of 2.75 days before testing with 82 percent sensitivity and 63 percent specificity. It is not a diagnostic test.
How long does Oura take to set a temperature baseline?
Oura says it calculates your normal temperature during the first couple of weeks of wearing the ring.
Does skin temperature change across the menstrual cycle?
Yes. In a 2019 study of 22 women, nightly finger skin temperature was 0.30 degrees C higher in the luteal phase than the follicular phase.
Sources
- Body Temperature, Oura Help Center
- WHOOP 101, WHOOP developer documentation
- Feasibility of continuous fever monitoring using wearable devices (Smarr et al.), Scientific Reports, 2020
- Detection of COVID-19 using multimodal data from a wearable device: results from the first TemPredict Study (Mason et al.), Scientific Reports, 2022
- Nocturnal finger skin temperature in menstrual cycle tracking: ambulatory pilot study using a wearable Oura ring (Maijala et al.), BMC Women's Health, 2019
