Buspirone is effective in treating clinical anxiety but, unlike classical anxiolytics, does not have anti-convulsant, sedative or muscle relaxant side-effects and does not interact with GABA. Buspirone may also differ from classical anxiolytics in requiring a period of 2 weeks or more to achieve its full therapeutic action. It has previously been shown that all anxiolytic drugs, including buspirone, reduce the frequency of reticular-elicited hippocampal rhythmical slow activity (RSA). The present experiments tested whether the time course of the effect of buspirone on rhythmical slow activity differed from that of the anxiolytic benzodiazepine chlordiazepoxide. Rats, implanted with reticular stimulation electrodes and subicular recording electrodes, received three intraperitoneal injections per day of buspirone (2.5 mg/kg), chlordizepoxide (5 mg/kg) or saline for 45 days. Both buspirone and chlordiazepoxide reduced the frequency of rhythmical slow activity on the first day of testing and Ro15-1788 (10 mg/kg) blocked the effects of chlordiazepoxide but not buspirone. There was no increase in the effect of buspirone with time. These results showed that, if the effect of anxiolytic drugs on rhythmical slow activity provides any basis for their clinical action, then some additional factors are required to explain both the delayed action of buspirone and the immediate action of classical anxiolytic drugs.