Reading your sleep score
Oura's 80 isn't Garmin's 80, and neither is Fitbit's. Each brand builds its sleep score differently. This guide compares what each manufacturer says goes into its score, how accurate wearables are, which metrics matter more than the headline number, and how to troubleshoot a low score without worrying about it.
What goes into each brand's score
All three combine sleep duration with some measure of quality, using proprietary formulas. Only the inputs are published, not the weights.
| Oura Sleep Score | Garmin sleep score | Fitbit / Google sleep score | |
|---|---|---|---|
| What the manufacturer says it uses | Contributors including total sleep, efficiency and latency, based on resting heart rate, body temperature, movement and estimated stages [1] | Duration plus quality: estimated stages, awake time, restlessness and HRV-based stress, using Firstbeat analytics [2] | Duration, time to sound sleep, sound sleep, restlessness, interruptions and full awakenings [3] |
| Scale | 0–100; 85+ "optimal", 70–84 "good" [1] | 0–100 [2] | Excellent 90–100, Good 80–89, Fair 60–79, Poor below 60; Google says most users average 72–83 [3] |
| Which sleep counts | One main sleep of at least 3 hours between 6 PM and 6 PM; naps are taken into account [1] | Not described in this source | Primary sleep window; needs sleep stages to produce a score [3] |
| Personalisation | Not described in this source | Not described in this source | Targets tailored to age, gender and time asleep or trying to sleep [3] |
Because the inputs and weights differ, scores aren't comparable across brands, and a change of device resets your trend.
How accurate are wearables?
Against laboratory polysomnography, consumer trackers generally do well at telling sleep from wake and estimating total sleep time, and are less accurate for individual stages. That was the finding in a 2021 study of seven devices in 34 healthy young adults (devices from around 2019) [4] and in a study of an earlier-generation Oura ring in 41 healthy adolescents and young adults [5]. Newer devices and algorithms may perform differently, and results may not carry over to older adults or people with sleep disorders. Treat "deep" and "REM" minutes as estimates and focus on trends.
Keep three questions apart: total sleep time accuracy (fairly good in these studies), stage accuracy (weaker), and score validity. The studies above checked sleep measurements, not whether a brand's overall score reflects sleep quality, and we found no independent validation of the composite scores themselves.
4 numbers that are often more useful than the score
- Total sleep time: a consensus statement recommends 7 or more hours a night regularly for adults [6]. Oura calls total sleep its most significant contributor [1]. Repeatedly short nights add up (see Sleep debt).
- Sleep efficiency (percent of time in bed spent asleep): low efficiency suggests fragmented sleep or a long time to fall asleep.
- Regularity: consistent bed and wake times are a core sleep-hygiene recommendation [7], and Oura includes sleep regularity as a Readiness contributor [8].
- Overnight resting heart rate and HRV: these aren't direct measures of sleep quality, but they often react to alcohol, late meals and illness [8][9], so they help explain a bad night.
Things that commonly hurt sleep (with evidence)
- Alcohol near bedtime: may help you fall asleep faster, but disrupts sleep in the second half of the night [10]. In a large observational study, it raised overnight heart rate and lowered HRV [9].
- Caffeine late in the day: in a small controlled study, 400 mg of caffeine taken 6 hours before bed still reduced total sleep [11].
- A warm bedroom: a review found heat exposure during the night increases wakefulness and reduces deep and REM sleep [12].
- Irregular timing: shifting bed and wake times works against your body clock [7].
- Late heavy meals or late exercise: Oura notes these can raise nighttime heart rate and delay its lowest point [8].
Troubleshooting a low score (illustrative examples)
Both scenarios are invented, not real user data.
Example 1: 8 hours, low score. Total sleep is normal, but overnight heart rate is higher and HRV lower for three nights, each after a late dinner with wine. The duration part of the score is fine; the physiological signals pulled it down. Moving dinner earlier and skipping weeknight alcohol is a more useful response than going to bed earlier.
Example 2: score fell after a new device. Someone switches from a watch to a ring and their average drops from the mid-80s to the mid-70s with no change in how they feel. Different inputs and weights mean the scores aren't comparable. Start a fresh baseline on the new device for a few weeks before judging.
Stop chasing the score
One bad night means little. A rolling average over 5 to 7 days tells you whether your routine is working. If checking your score makes you anxious or makes it harder to sleep, take a break from it. How rested you feel during the day matters more than the number.
When to talk to a doctor
Loud snoring, pauses in breathing, gasping at night, heavy daytime sleepiness or long-lasting insomnia are reasons to see a doctor. Wearables can't diagnose sleep disorders.
How RadiantHealth uses your sleep data
RadiantHealth reads sleep from one connected source per category (Garmin, Oura or Health Connect; you can change which). It uses sleep duration, a simplified 7-day sleep-debt estimate against a fixed 7.5-hour reference, and your recent trend as inputs to the morning briefing, together with HRV and training load. It doesn't create its own sleep score.
FAQ
What is a good sleep score?
Each brand defines its own scale. Oura, for example, calls 85–100 optimal and 70–84 good. Your total sleep, regularity and trend are more useful than any single number.
Is deep sleep tracking accurate?
Less so than total sleep time. Validation studies show consumer devices are less accurate at stages, so treat stage data as an estimate.
Why is my sleep score low when I slept 8 hours?
Scores also weigh things like stages, restlessness, awakenings and, depending on the brand, heart rate, temperature or HRV-based stress. Alcohol, a late meal or illness can lower the score even after a long night.
Sources
- Oura Member Care: Sleep Score (last updated 25 Sep 2026; accessed 23 Oct 2026).
- Garmin blog: How Garmin watches track your sleep and calculate sleep score.
- Google Help (Fitbit / Google Health): Sleep score.
- Chinoy ED et al. Performance of seven consumer sleep-tracking devices compared with polysomnography. Sleep 2021;44(5):zsaa291. doi:10.1093/sleep/zsaa291
- de Zambotti M et al. The sleep of the ring: comparison of the ŌURA sleep tracker against polysomnography. Behav Sleep Med 2019;17(2):124-136. doi:10.1080/15402002.2017.1300587
- Watson NF et al. Recommended amount of sleep for a healthy adult. Sleep 2015;38(6):843-844. doi:10.5665/sleep.4716
- Irish LA et al. The role of sleep hygiene in promoting public health. Sleep Med Rev 2015;22:23-36. doi:10.1016/j.smrv.2014.10.001
- Oura Member Care: Readiness Contributors.
- Pietilä J et al. Acute effect of alcohol intake on cardiovascular autonomic regulation during the first hours of sleep. JMIR Ment Health 2018;5(1):e23. doi:10.2196/mental.9519
- Ebrahim IO et al. Alcohol and sleep I: effects on normal sleep. Alcohol Clin Exp Res 2013;37(4):539-549. doi:10.1111/acer.12006
- Drake C et al. Caffeine effects on sleep taken 0, 3, or 6 hours before going to bed. J Clin Sleep Med 2013;9(11):1195-1200. doi:10.5664/jcsm.3170
- Okamoto-Mizuno K, Mizuno K. Effects of thermal environment on sleep and circadian rhythm. J Physiol Anthropol 2012;31:14. doi:10.1186/1880-6805-31-14
Get a sleep-aware daily plan
Free plan available. Connect your tracker for a morning briefing that takes last night's sleep into account.
Start →