Abstract Background Circadian theories for major depressive disorder have suggested that the rhythm of the circadian pacemaker is misaligned. Stable phase relationships between internal rhythms, such as temperature and rest/activity, and the external day-night cycle, are considered to be crucial for adapting to life in the external environmental. Therefore, the relationship and possible alterations among (i) light exposure, (ii) activity rhythm, and (iii) temperature rhythm could be important factors in clinical depression. This study aimed to investigate the rhythmic alterations in depression and evaluate the ability of chronobiological parameters to discriminate between healthy subjects and depressed patients. Methods Thirty female subjects, including healthy subjects, depressed patients in the first episode, and major recurrent depression patients. Symptoms were assessed using Hamilton Depression Scale, Beck Depression Inventory and Montgomery-Äsberg Scale. Motor activity, temperature, and light values were determined for 7 days by actigraph, and circadian rhythms were calculated. Results Depressed groups showed a lower amplitude in the circadian rhythm of activity and light exposure, but a higher amplitude in the rhythm of peripheral temperature. The correlation between temperature and activity values was different in the day and night among the control and depressed groups. For the same level of activity, depressed patients had lowest temperature values during the day. The amplitudes of temperature and activity were the highest discriminant parameters. Conclusions These results indicate that the study of rhythms is useful for diagnosis and therapy for depressive mood disorders.