Buspirone, a 5-HT1A agonist, has been shown to decrease the intake of ethanol when given as a single dose to rats with a psychological dependence induced according to our rat model of alcoholism. The present experiment evaluates the effects different treatments with buspirone have on voluntary ethanol intake in these psychologically dependent rats. As a first treatment, buspirone was given once daily for 23 days at the dose of 20 mg/kg/day. Ethanol was withheld except for the first and the last day of the treatment. On the first day, the buspirone injection decreased ethanol intake from the pretreatment value (1.94+/-0.18 g/kg/day), down to 1.36+/-0.18 g/kg (p < 0.01, n = 12). The rats were again given a choice between water and 10% ethanol after the last injection of buspirone. During the following 24 hr period, the ethanol intake was increased to 3.56+/-0.24 g/kg/day (p < 0.001 vs. the pretreatment intake, n = 12). A loss of correlation with the pretreatment intake of ethanol indicated an altered regulation of ethanol intake for approximately 3 more weeks. Fifteen weeks after the start of the first treatment, buspirone (20 mg/kg) was re-tested as a single dose, with no effect on ethanol intake. Twenty-two weeks after the start of the first treatment, a 1-week treatment with 20 mg/kg/day of buspirone was started. During this treatment, the rats had a continuous choice between 10% ethanol and water. There was, as in the first re-test, no effect on ethanol intake on the first day of the treatment. However, on the last 2 days of the treatment, the ethanol intake was increased to 2.86+/-0.28 g/kg and to 2.89+/-0.26 g/kg respectively (p < 0.05, n = 10 on both days, compared with the pretreatment intake of 1.78+/-0.36 g/kg). Thus, an acute dose of buspirone can decrease voluntary ethanol intake in psychologically dependent rats, but long-lasting changes in the effect of buspirone seem to develop during a 3-week treatment period.