PURPOSE: Sleep quality and quantity are associated with an increased risk for chronic diseases, but many studies rely on self-report logs to assess sleep. Wearable devices estimate sleep quantity and quality, but it is not clear how these device estimates of sleep compare to self-report logs. The purpose of this study was to 1) compare sleep estimates collected by the devices and the self-report logs and 2) determine if estimates of sleep duration are different between individuals classified as “good” and “poor” sleepers. METHODS: In this cross-sectional observational study, participants (n=26, average age=30, 16 females) simultaneously wore devices on the wrist and thigh for seven consecutive days and tracked their wake and sleep times using a daily sleep log (SL). At the end of the 7-days, they completed the Pittsburgh Sleep Quality Index (PSQI) questionnaire, which classified their sleep quality as “good” or “poor.” Repeated measures ANOVA and Pearson correlations were used to compare average sleep duration across monitors and an independent t-test was used to compare the sleep duration estimates between “good” and “poor” sleepers. RESULTS: There was a significant difference between the sleep duration estimates collected by the sensors (THIGH=557.8±78.5 min; WRIST=564.4±53.2 min) and the self-report logs (SL=492.2±43 min; PSQI=472±55.2 min) (p<0.0001). There was a strong significant positive correlation of 0.6 between the sleep duration estimates collected by the PSQI and the sleep log. Between “good” and “poor” sleepers, there was no significant difference in sleep pattern estimates. CONCLUSIONS: Sleep duration estimates collected by the sensors tended to be higher than self-report estimates. This difference may be due to bias in self-reporting sleep, or misclassification of lying time (e.g., watching TV) that is classified as sleep time by devices. Sleep duration estimates did not differ between those with “good” or “poor” sleep quality, highlighting that sleep is a multidimensional physiological construct. To ensure coherent translation in sleep studies, there is a need to develop standardized methods that will produce comparable estimates of sleep across device and self-report data.