Despite the long history in medicine, the pathophysiological mechanism(s) of seasonal affective disorder (SAD) remain largely unknown. By employing a meta-analytic methodology, the authors of this study attempted to verify the validity of different pathophysiological mechanism(s) proposed for SAD. The findings showed that for phototherapy of medium light intensity, a combination of morning–evening therapy regime yielded the best therapeutic effect, and the antidepressant effect of the morning–evening light regime was superior to a single pulse of light administered at other times of day. Furthermore, the data showed that the antidepressant effect of a single pulse of light was similar for morning, midday, and evening light. These findings supported the photon-count hypothesis and refuted the proposed photoperiod, melatonin, and phase-shifting models of SAD. Copyright © 1997 Elsevier Science B.V. All rights reserved.