The involvement of chronobiological mechanisms in the antidepressant response to such non-drug treatments as bright light, physical exercise and sleep deprivation still remain to be clarified. We compare the efficacy of several treatment strategies for seasonal and non-seasonal depression and discuss possible the contribution of chronobiological and psychological mechanisms in antidepressant response. The therapeutic effects were tested at the medical academic hospital near Novosibirsk (55 degrees North) in 138 subjects, either with winter depression or with non-seasonal depression or without depression (n = 41, 64 and 33, respectively). One-week monotreatments were either 2-hour 2500 lux cool-white incandescent light from 14:00 (n = 9, 9, 9, respectively) or 1-hour physical exercise from 13:00 (n = 9, 9, 9, respectively). One-week combined treatments included a night of total sleep deprivation followed by either 2-hour bright light from 14:00 (n = 8, 12, 0, respectively) or 1-hour physical exercise either under ordinary room light from 13:00 (n = 0, 12, 0, respectively) or under bright light from 12:00 (n = 5, 11, 0, respectively). The results indicate that, in subjects left without antidepressant treatment for a week (n = 10, 11, and 15, respectively), the 21-item Hamilton Depression Rating Scale score did not change significantly. The beneficial effects of total sleep deprivation were similar in seasonal and non-seasonal depression. The seasonals exhibited better response to bright light compared to non-seasonals. After sleep deprivation the substantial further improvements were produced by either lighting or exercising. Compared to the patients exercising under ordinary room light, the patients exercising under bright light did not gain an additional benefit. In general, winter depression was well-treated with either exercise or light, while the most promising treatment for non-seasonal depression was physical exercise combined with sleep deprivation. Bright light or physical exercise administered in the middle of the day were not less favorable compared to the treatments in the morning hours, although it is unlikely that they considerably challenged patient's chronobiology. It was concluded that the placebo effect would account for a large portion of clinical response to open non-pharmacological treatments. Therapeutic hops and visibility of such treatments would explain their high antidepressant efficacy in comparison with pharmacological trials applying a double blind cross-over design. In particular, the excellent response of patients with winter depression to light therapy might be related to their tendency to attribute a high symbolic value to bright light and associate their bad mood with a dark season.