In patients with winter seasonal affective disorder (SAD) the onset of a depressive episode is probably associated with the decreasing amount of light during the autumn. A highly predictable onset of a recurrent depressive episode with seasonal pattern provides a rationale for testing the efficacy of bright-light treatment as a preventive measure. Twelve out-patients with winter SAD were assigned to start bright light treatment either when they were well, or not to start it until the first symptoms of depression had already emerged. The severity of depressive symptoms was prospectively rated for a parallel randomized comparison. Bright light given well in advance of the emerging symptoms prevented a depressive episode. Clinical remission was significantly more frequent in the former subgroup of the patients in January and in March. To sum up, bright light can be successfully administered as prophylactic treatment for the prevention of winter SAD.