Long-term (33-35 days) castration caused a significant increase in the duration of immobility of male and female mice in the tail suspension test (an animal model of depression), and a significant decrease in the maximum number (Bmax) of [3H]imipramine binding sites in the cerebral cortex of male mice. In the tail suspension test, gonadotropin-releasing hormone (GnRH), s.c. injected 3 times at 3-h intervals at doses of 0.2, 2 or 20 micrograms/kg, did not significantly modify the duration of immobility of castrated animals and did not reduce that of sham-operated ones, while desipramine (20 mg/kg s.c. 1 h before testing) restored immobility to normal in castrated animals and reduced it significantly in sham-operated ones. The same treatment schedule with GnRH produced an increase in the number of [3H]imipramine Bmax in cortical membranes that was statistically significant at the dose of 2 micrograms/kg. It is concluded that the castration-induced depression-like behavior in mice seems not to be due to the decreased levels and release of GnRH, and that GnRH has no antidepressant-like effect in mice, at least at our dose levels; however, GnRH seems to increase the number of cortical [3H]imipramine binding sites.