Recent experiments in primates indicate that the phenothiazine drug chlorpromazine hydrochloride inhibits both bile salt-dependent and -independent bile flow in a predictable fashion. Because a significant fraction of bile salt-independent bile flow is postulated to depend upon the activity of a canalicular membrane Na+,K+-ATPase, we have examined the effects chlorpromazine hydrochloride and its metabolites on the ATPase activities of canalicular-enriched rat liver plasma membranes. Chlorpromazine inhibited the activities of both Mg2+- and Na+,K+-ATPases with a linear dose-response relationship between 10 and 100 micronM. The inhibition of Na+,K+-ATPase was pH dependent, showing a maximal inhibition at pH 7.8. Over the pH range 7.0 to 8.2, the inhibition was significantly reduced with the addition of glutathione and was augmented under experimental conditions (ultraviolet irradiation and peroxidase-H2O2) that promote the formation of the chlorpromazine semiquinone free radical. The 7-hydroxychlorpromazine metabolite was as active an inhibitor as the parent drug; however, two sulfoxide metabolites, chlorpromazine sulfoxide and 7-hydroxychlorpromazine sulfoxide, were less effective inhibitors of Na+, K+-ATPase. Our data are consistent with the hypothesis that chlorpromazine cholestasis may be a result of a direct toxic effect on the ATPase activities of hepatic canalicular membranes. Our results further suggest that if chlorpromazine cholestasis occurs through an interaction with canalicular membrane ATPases, the degree of cholestasis may well be influenced by the extent of the conversion of the drug to its more active (free radical) or minimally active (sulfoxide) metabolites and by the local environment (pH and glutathione concentration) of the canalicular membrane.