OBJECTIVE: Sleep disturbance, a core feature of bipolar disorder, is associated with residual mood symptoms, mood episode recurrence and suicide ideation. We investigated the effect of evening light exposure on sleep in patients with bipolar disorder. METHODS: In this longitudinal analysis, we measured the sleep parameters of 207 outpatients with bipolar disorder using actigraphy at their homes for seven consecutive nights. We measured the white-light illuminance and the irradiance of each wavelength during the 4 hours before each participant's bedtime. We used mixed-effect linear regression analysis for repeated measures to evaluate the effect of evening light exposure on subsequent sleep parameters. RESULTS: The median white-light illuminance was 25.8 lux (interquartile range, 12.9-50.1 lux). In a multivariable model adjusted for potential confounders, we found higher white-light illuminance to be significantly associated with lower sleep efficiency (per log lux: 95% confidence interval = [-1.328, -0.133]; p = 0.017), prolonged sleep-onset latency (95% confidence interval = [0.006, 0.172]; p = 0.035) and longer wake after sleep onset (95% confidence interval = [1.104, 4.459]; p = 0.001). This effect size was larger in the younger age group (aged < 44 years) stratified by median age. Higher irradiance of the blue wavelength range was significantly associated with longer wake after sleep onset, a result similar to those for the green and red wavelength ranges. CONCLUSION: We observed significant associations between evening light exposure and subsequent sleep in patients with bipolar disorder. The effects of various light wavelengths on sleep in bipolar disorder require further investigation.