Background. Several studies have described but not formally tested discrepancies between subjective and objective measures of sleep. Study Objectives. To test the hypothesis that patients with bipolar disorder display a systematic bias to underestimate sleep duration and overestimate sleep latency. Methods. Actimetry was used to assess sleep latency and duration in 49 euthymic participants (bipolar = 21; healthy controls = 28) for 5–7 days. Participants simultaneously recorded estimated sleep duration and sleep latency on a daily basis via an online sleep diary. Group differences in the discrepancy between subjective and objective parameters were calculated using t-tests and corrected for multiple comparisons. Results. Patients with bipolar disorder significantly underestimated their sleep duration but did not overestimate their sleep latency compared to healthy controls. Conclusions. Studies utilizing diaries or questionnaires alone in patients with bipolar disorders may systematically underestimate sleep duration compared to healthy controls. The additional use of objective assessment methods such as actimetry is advisable.