Objectives: Theoretical accounts and psychological interventions for bipolar disorder indicate that disruption of circadian rhythms is important, both in affective episodes and as a vulnerability factor in subsyndromal periods. This study aims at assessing both circadian activity and sleep patterns using actigraphy within a bipolar sample experiencing low levels of subsyndromal symptoms. It is hypothesized that such participants will display circadian activity disruption in spite of low levels of symptoms. Methods: This study employed a mixed design with cross-sectional assessment of mood and week-long (7-day) recording of actigraphy data. All clinical participants were psychiatric outpatients within a UK NHS Hospital. Nineteen bipolar patients and 19 age- and gender-matched controls wore an actigraph for 7 days to obtain sleep and circadian activity data. SCID was used to confirm DSM-IV diagnostic status. Self-report measures of mood were obtained from both groups. Results: Bipolar patients were found to have less stable and more variable circadian activity patterns than controls. Regression analysis indicated that variability alone was a significant independent predictor of diagnostic group. There was evidence from raw activity data that bipolar patients were also less active than controls. These differences were not associated with levels of subsyndromal symptoms. Bipolar patients did not differ from controls on any of the sleep indices used. Conclusions: Circadian activity disruption is apparent in bipolar patients even when not acutely ill. This finding is not associated with the presence of sleep disturbance. Should such patterns be replicated interventions to address both circadian instability and individual attributions for the effects of such instability are likely to be relevant to successful psychological interventions.