Eleven females and five males with fall/winter seasonal affective disorder were randomly assigned to 7-day treatment regimens from 8 p.m. to 10 p.m. using identical light at 2000 or 300 lux. A modified Hamilton Rating Scale for Depression and a Beck Depression Inventory were administered before treatment, after treatment # 7, and 2 weeks after phototherapy was terminated. Analysis of variance with repeated measures revealed a significant interaction between sex of the patient, intensity of the lights, and day of rating for scores on both the modified Hamilton Rating Scale for Depression and the Beck Depression Inventory. For both measures, the interaction occurred because all groups showed a decrease in depression ratings during the phototherapy exposure period, but only females at the higher intensity continued to have low depression scores 2 weeks after light treatment had stopped. These data indicate that bright light at both high (2000 lux) and low (300 lux) intensities is able to reduce depression in patients with seasonal affective disorder. The data also indicate that both sex of the patient and intensity of the light may interact to determine the latency to relapse.