We examined the efficacy of estrogen in the treatment of depression in perimenopausal women with and without hot flushes. Women with perimenopause-related depression were randomized in a double-blind parallel design to receive either 17beta-estradiol or placebo for 3 weeks. Subsequently, women receiving estradiol during the first 3 weeks continued receiving estradiol for an additional 3 weeks, whereas women who had received placebo crossed over to estradiol for 3 weeks. Outcome measures included standardized mood rating scales and a visual analog scale self-report instrument. Of 34 female subjects, 16 received estradiol first and 18 received placebo first. After 3 weeks of estradiol, standardized mood rating scale scores and visual analog scale symptom scores (eg, sadness, anhedonia, and social isolation) were significantly decreased compared with baseline scores (P <.01) and were significantly lower than scores in women receiving placebo (P <.01), who showed no significant improvement. Neither the presence of hot flushes nor the duration of treatment (3 weeks vs 6 weeks) influenced outcome. A full or partial therapeutic response was seen in 80% of subjects receiving estradiol and 22% of those receiving placebo. In this preliminary study estradiol replacement effectively treats perimenopausal depression independent of its salutary effects on vasomotor symptoms.