Drug-induced improvement of depression may be mediated by changes in sleep physiology. The aim of this study was to relate changes in sleep polygraphic variables to clinical state during treatment with citalopram, a highly specific serotonin uptake inhibitor. Sixteen patients took part. The study was single-blind and uncontrolled. A 1-week wash-out period was followed by 1 week of placebo administration, a medication period of 5 weeks, and a 1-week placebo period. For the entire group a significant decrease of rapid eye movement sleep (REMS) and a significant lengthening of REMS latency were observed initially as well as at the end of treatment. No changes in sleep continuity were found, but non-REMS stage 2 (percentage) was significantly increased. On the basis of clinical change, as expressed by the scores of the Hamilton Rating Scale for Depression, at the end of the citalopram treatment the patient group was split in two halves: eight less and eight more improved patients. The groups did not differ with respect to any sleep polygraphic variable.