A woman with seasonal affective disorder (SAD) remitted within a week in each of six separate trials of light therapy. She remitted within 2 weeks of initiating citalopram treatment. Light treatment in the morning advanced and improved sleep, whereas citalopram delayed sleep and induced intermittent awakenings. These opposite patterns suggest that sleep deprivation or sleep shifts were not crucial for eliciting therapeutic response. Light and citalopram both selectively reduced intake of sweet carbohydrate parallel with improvement, implicating CNS serotonergic mechanisms in the interaction of mood and food in winter.