Oura Ring Faces Proposed Class Action Over Sleep-Tracking Accuracy Claims

PROPOSED CLASS ACTION · COMPLAINT FILED AUG. 20, 2026

A federal complaint challenges how Oura markets sleep-stage accuracy. The allegations are serious, but they are not findings—and the science varies sharply by metric and population.

A proposed nationwide class action filed in federal court in San Francisco is challenging one of Oura’s most important selling points: the accuracy of the sleep-stage estimates produced by its smart rings.^[1]

The complaint, filed August 20, 2026, names Oura Inc. and Finland-based Oura Health Oy and identifies California resident Madison Surber as the named plaintiff. The case is Surber v. Oura Inc. et al., No. 3:26-cv-08686, in the U.S. District Court for the Northern District of California.^[1]

That wording matters. The source reviewed here is a complaint containing one side’s allegations and requests for relief, not a merits decision or class-certification order.^[1] Authoritative current-docket verification is required immediately before publication.

What the lawsuit alleges

Surber says she bought an Oura Ring 4 Gold for about $513.68 in May 2025 after seeing Oura’s website and retail marketing. She alleges she relied on promises that the device could provide accurate stage-by-stage sleep data, then concluded after using it that the ring did not reliably track how long she slept, when she awakened, or the quality of her sleep.^[1]

The complaint points to Oura phrases including “Built for accuracy,” “Unparalleled Accuracy,” a claim of 79% agreement with polysomnography for four-stage classification, and a current Oura Ring 5 product-page statement of “95% Sleep Staging Accuracy” “Compared to clinical sleep lab.”^[1][2]

The lawsuit’s central theory is that Oura’s rings do not directly measure the neurological signals used by a clinical polysomnography study—brain waves, eye movements, and muscle activity—yet Oura markets algorithmic estimates as if they were equivalent to direct clinical measurement.^[1]

The complaint calls those estimates inaccurate and misleading, argues that buyers paid a price premium, and says Oura knew or should have known the limits of a finger-worn device.^[1] Those are the plaintiff’s allegations, not established facts.

How large is the proposed class?

The complaint proposes two groups:^[1]

• A nationwide class of U.S. purchasers who bought covered Oura products for personal use within the applicable limitations periods.^[1]
• A California subclass of purchasers who bought covered products in California during the four years before the complaint was filed.^[1]

The filing describes the products as all U.S.-sold Oura sleep-tracker models and specifically names Oura Ring 5, Oura Ring 4, and Oura Ring 4 Ceramic. It asserts that the nationwide group has tens of thousands of members or more and that the California group has thousands or more.^[1]

Those are the definitions the plaintiff asks the court to certify.^[1] The complaint itself does not show that the request was granted; authoritative docket verification is required before describing the current certification posture.

Seven legal claims—and what the plaintiff wants

The 53-page complaint pleads seven counts:^[1]

1. Fraud by misrepresentation
2. Unjust enrichment
3. Violation of California’s Unfair Competition Law
4. Violation of California’s False Advertising Law
5. Violation of California’s Consumers Legal Remedies Act
6. Breach of express warranty
7. Breach of implied warranty under the Song-Beverly Act^[1]

The plaintiff asks for class certification, declaratory and injunctive relief, changes to Oura’s advertising practices, damages, restitution or disgorgement where legally available, punitive damages or penalties where permitted, attorneys’ fees, costs, and interest. The complaint also demands a jury trial.^[1]

The complaint requests monetary relief but does not itself award money or create a claims process.^[1] Readers should not treat a newly filed complaint as a court-approved refund notice.

The strongest factual point in the complaint

The clearest undisputed distinction is between direct measurement and algorithmic estimation.^[3][4]

Oura itself explains that clinical polysomnography uses EEG to measure electrical brain activity, EOG to measure eye movement, EMG to measure muscle activity, and other channels. Oura also states that using its ring is “not the same as a PSG” because the ring does not measure electrical brain activity or eye movements.^[4]

Instead, Oura’s sleep algorithm uses signals available at the finger—including movement, temperature, heart rate, and heart-rate variability—to infer likely sleep stages. Oura says this multichannel approach achieved 79% agreement with PSG for four-stage classification in the research underlying its Sleep Staging Algorithm 2.0.^[3]

The current Oura Ring 5 sales page goes further, displaying “95% Sleep Staging Accuracy” beside “Compared to clinical sleep lab.” The visible product page does not explain on the face of that statistic which study, cohort, staging metric, aggregation rule, or Ring 5 validation protocol produces the 95% number.^[2]

That disclosure gap may become important.^[2] “Agreement,” “overall accuracy,” “sensitivity,” “specificity,” “reliability,” and agreement in nightly summary totals are different statistical concepts.^[3][5][7] They cannot safely be placed on one scale without the underlying protocol.^[3][5][7]

Why “the ring cannot measure brain waves” does not end the scientific debate

The complaint argues that because a ring lacks EEG, EOG, and EMG, it cannot accurately measure sleep stages.^[1] That claim is broader than the available research supports.^[5][7]

A device can estimate a hidden state from correlated signals without directly sensing the defining physiological channel.^[3][7] Whether the estimate is useful depends on the algorithm, population, metric, and intended use.^[5][6][7] Finger-derived movement and cardiovascular patterns contain information that Oura’s algorithm uses to estimate sleep and wake states.^[3] The legal question is not simply whether Oura has scalp electrodes; it is whether Oura’s particular advertising conveys a level and type of accuracy that the evidence substantiates for reasonable buyers.^[1][2]

Oura’s own language reflects both sides. It emphasizes high agreement with PSG while acknowledging that the ring is not a PSG and does not measure brain or eye activity.^[3][4] The lawsuit says the high-visibility promise overwhelms the lower-visibility limitation.^[1] This packet does not characterize Oura’s litigation position because authoritative docket coverage was unavailable.

What the studies actually show

Oura-backed multi-night study: strong results in generally healthy adults

A 2024 study of Oura Ring Gen3 and Oura Sleep Staging Algorithm 2.0 evaluated 96 generally healthy Japanese adults, contributing 421,045 thirty-second epochs over as many as three PSG nights each. It reported no statistically significant difference from PSG for multiple nightly measures, including total sleep time and time in light and deep sleep. It reported sleep/wake accuracy of about 91.7% to 91.8%, and stage-specific accuracy ranging from 75.5% for light sleep to 90.6% for REM sleep.^[7]

That study provides meaningful evidence that algorithmic estimation can perform well under its protocol. But its conflict-of-interest statement says Oura Health Oy provided financial support to several authors.^[7] Funding does not invalidate a study, but it belongs in any fair account—especially when independent studies find weaker performance in different populations.

Clinical sleep-lab patients: averages can hide large individual errors

A 2025 Scientific Reports study compared Oura Gen3, SleepOn, and Circul rings against diagnostic PSG in sleep-lab patients. Oura’s binary sleep/wake accuracy averaged 85.03%, but wake sensitivity was only 46.19%. For four-stage classification, the Oura confusion matrix reported sensitivities of roughly 46% for wake, 56% for light sleep, 50% for deep sleep, and 55% for REM.^[5]

The paper also found wide person-to-person differences. Oura overestimated total sleep time by 11.74 minutes on average, but individual differences ranged from 54 minutes under PSG to 97.5 minutes over it. It overestimated REM by 31.56 minutes on average; individual REM differences ranged from 40 minutes under to 172 minutes over PSG. The authors warned that reasonable group averages could mask substantial individual errors and limited transfer from healthy young samples to clinical patients.^[5]

The study had its own limits: only 31 usable Oura nights remained after dropouts, participants wore multiple rings simultaneously, the tested device was Gen3, and the protocol did not include a long adaptation period.^[5]

Older adults: substantially weaker performance in a small one-night study

A 2026 study compared consumer sleep devices with at-home PSG in 13 younger adults and 19 older adults. In the older group, Oura underestimated total sleep time by about 75.5 minutes on the PSG night. The authors concluded that older adults—and clinicians or researchers considering consumer devices as PSG replacements—should use caution.^[6]

That result matters because algorithms trained or validated in healthier and younger cohorts may not generalize to older adults or people with fragmented sleep. It does not establish that every Oura user receives a 75-minute error, and the sample was small and based on one PSG night.^[6]

The “coin toss” headline needs qualification

The complaint describes Oura’s real-world stage estimates as having roughly “a coin flip’s chance” of being correct. That line reflects its reading of independent clinical data and is already driving headlines.^[1][8]

It should not be presented as a universal measured truth. The clinical study’s four-stage sensitivities clustered around 46% to 56% in its patient sample, but other studies, methods, and populations report higher results.^[5][7] Even within one paper, binary sleep/wake accuracy can be high while wake detection or precise stage classification remains weak.

The scientifically honest conclusion is narrower: Oura’s performance is not one fixed percentage.^[3][5][7] Results vary substantially by metric, sleep stage, participant population, device generation, algorithm version, and whether the question concerns group averages or a specific individual.^[5][6][7]

Why this matters beyond one company

Sleep tracking can change behavior. An American Academy of Sleep Medicine survey found that 68% of people who had tried a tracker changed behavior based on what they learned.^[9] A later AASM survey found that worries about sleep problems had cost many adults sleep and recommended that users avoid obsessing over nightly data.^[10]

That makes calibration and communication important.^[9][10] A wellness device can be useful for observing personal trends without being suitable for diagnosis or precise clinical staging.^[4][10] Trouble begins when a consumer interprets a nightly score as a medical verdict, changes treatment, ignores symptoms, or becomes anxious because an algorithm disagrees with how they feel.^[9][10]

What Oura users should do now

Current owners do not need to throw away their rings. A cautious approach, consistent with AASM guidance and the device/PSG distinction, is to:^[4][9][10]

• Treat sleep scores and stage estimates as wellness trends, not diagnoses.^[4][10]
• Look for multi-night patterns rather than overreacting to one night.^[4][10]
• Pay attention to how rested you feel and whether symptoms persist.^[9]
• Discuss loud snoring, gasping, severe daytime sleepiness, insomnia, or other concerning symptoms with a clinician rather than relying on a wearable.^[9][10]
• Avoid changing medication, CPAP settings, or medical care solely because of ring data.^[4][10]
• Preserve receipts, screenshots of claims shown at purchase, and subscription records if you believe you were misled—but do not send personal health records to strangers or unverified “claims” websites.
• Remember that the complaint itself is not a class notice, settlement claim form, or court-approved refund process.^[1] Check an authoritative docket or court-approved notice before submitting information anywhere.

What happens next

The immediate questions are procedural: service, Oura’s response, possible motions to dismiss, and later arguments over class certification.^[1] Oura may challenge standing, nationwide application of California law, the proposed class definitions, the interpretation of its advertising, causation, reliance, damages, or the scientific premise of the complaint.^[1]

TechCrunch reported that Oura did not immediately respond to its request for comment.^[8] That deadline-specific statement is not a refusal to respond and should not be treated as an admission. Oura’s current litigation position requires authoritative docket or company-source verification.

The case deserves close attention because it puts a familiar consumer-tech problem into unusually sharp focus: how should a company describe an algorithm that produces useful estimates for some people while sometimes diverging substantially from a clinical reference for others?

The answer will not come from one dramatic percentage.^[2][3][5] It will come from the wording consumers saw, the scientific protocol behind each number, the populations on which the models were validated, and what Oura disclosed about the difference between a trend-estimation tool and a clinical sleep study.^[1][4][6]

Editorial note: This article reports allegations and requested relief in a complaint.^[1] It does not establish Oura’s liability, current class-certification posture, or anyone’s current eligibility for compensation; those require authoritative current-docket verification.

Frequently asked questions

Sources

1. Surber v. Oura class action complaint
   

   “Nationwide Class: All purchasers who, within the applicable statute of limitations periods, purchased the Products in the United States for purposes other than resale”

   “California Subclass: All purchasers who, within four years prior to the filing of this Complaint, purchased the Products in California for purposes other than resale”

2. Oura Ring 5 product page
   

   “Compared to clinical sleep lab”

   “95%Sleep Staging Accuracy”

3. Oura Sleep Staging Algorithm 2.0
   

   “By leveraging every signal from the Oura Ring (movement, temperature, heart rate, and HRV), the Sensors study found Oura’s sleep staging algorithm achieves 79% agreement with gold-standard polysomnography (PSG) for classifying the four stages of sleep (wake, light, deep, and rapid eye movement (REM) sleep).”

4. Oura explanation of PSG and ring differences
   

   “While using an Oura Ring is not the same as a PSG (as it does not measure electrical brain activity or eye movements), research shows that Oura’s latest sleep staging algorithm achieves 79% agreement with gold-standard polysomnography (PSG) for 4-stage sleep classification.”

5. Herberger et al. 2025 clinical ring tracker performance
   

   “Compared to TST values computed from PSG, the Oura ring overestimated TST by 11.74 min on average (SD = 37.11 min.), and individual-level differences ranged from − 54.00 min to 97.50 min.”

   “In summary, our findings reveal limited agreement between previous ring tracker evaluations in healthy, young individuals and our sleep-lab patient sample.”

6. Searles et al. 2026 young and older adults
   

   “Older adults interested in tracking their sleep and clinicians and researchers using consumer devices as a replacement for polysomnography should use caution when interpreting results from these devices.”

   “In older adults, devices underestimated total sleep time (Fitbit bias = -74.5 minutes, p=.012; Oura bias = -75.5, p = <0.0001; Withings bias = -45.7, p=.083; Sleep Score Max bias = -56.5, p=.001)”

7. Svensson et al. 2024 Oura Gen3 validation
   

   “The Oura Ring did not significantly differ from PSG for the measures time in bed, total sleep time, sleep onset latency, sleep period time, wake after sleep onset, time spent in light sleep, and time spent in deep sleep.”

   “Conflict of interest statement: Declaration of competing interest Thomas Svensson reports financial support was provided by Oura Health Oy.”

8. TechCrunch Oura lawsuit report
   

   “Oura did not immediately respond to TechCrunch’s request for comment.”

9. AASM sleep tracker survey
   

   “According to a 2023 survey from the American Academy of Sleep Medicine, more than one-third of Americans (35%) have used an electronic sleep-tracking device. A majority of those who have tried one found the sleep tracker was helpful (77%), and many have changed behavior because of what they learned (68%).”

10. AASM sleep tracking and orthosomnia survey
    

    “According to the survey, most adults (76%) have lost sleep due to worries about sleep problems.”

    “Only check data in the morning — never at night, and don’t obsess over it.”

Read Oura’s explanation of PSG and ring estimates

Reporting cutoff: August 22, 2026 at 4:27 p.m. EDT / 20:27 UTC. Current post-complaint docket posture was not independently verified. Company and product names belong to their owners; The IT Guys is not affiliated with the parties or cited publishers.