Background: Disruption of circadian rhythmicity is associated with increased vulnerability to mood disorders. Previous studies have used subjective reports of activity patterns but the availability of accelerometer-based data in large numbers of participants of the UK Biobank cohort permits the derivation and analysis of new, objectively-ascertained rhythmicity parameters. Seasonal factors have been linked to circadian disruption and increased depressive symptoms, but previous studies have rarely accounted for effects of lifestyle and environmental factors. Using extensive health and lifestyle data from the UK Biobank, we tested for associations between objectively-assessed circadian rhythmicity and mental health and wellbeing phenotypes, including history of mood disorder. We then examined evidence of seasonal patterns in recent depressive symptoms. Methods: Using wrist-worn accelerometry data from 91,105 UK Biobank participants, a circadian relative amplitude (RA) variable was derived, reflecting the distinction between the most and least active periods of the day. Cross-sectional associations between low RA and mood disorder, wellbeing and cognitive variables were examined. To test for seasonality of depressive symptoms, we fitted cosinor models to a total depressive symptoms score and scores of low mood, anhedonia, tenseness and tiredness scores in >150,000 participants. Associations of depressive symptoms with day length were then examined. Findings: A quintile reduction in RA was associated with increased risk of lifetime major depressive disorder (odds ratio (OR) = 1.06, 95% CI 1.04, 1.08) and bipolar disorder (OR = 1.11, 95% CI 1.03, 1.20), as well as with greater mood instability, higher neuroticism, more subjective loneliness, lower happiness and health satisfaction, and slower reaction time. The associations were independent of demographic, lifestyle, education and overall activity confounders. Seasonality of total depressive symptom scores, anhedonia and tiredness scores was observed in women but not men, with peaks in winter. In women, increased day length was associated with reduced low mood (incidence rate ratio (IRR) = 0.997, SE = 0.001) and anhedonia scores (IRR = 0.996, SE = 0.001), but longer day length was associated with increased tiredness (IRR = 1.004, SE = 0.001). Interpretation: This large, population-based study provides evidence that circadian disruption is associated with a range of adverse mental health and wellbeing outcomes. We also provide evidence of seasonal variation in depressive symptoms in women, with association between shorter days and increased feelings of low mood and anhedonia. The depth and breadth of physical and mental health, sociodemographic, lifestyle, environmental and activity data in UK Biobank affords the opportunity to examine chronobiological contributions to mood disorder on an unprecedented scale.