We have previously studied the pharmacokinetics of the cyclosporine-ketoconazole interaction and shown that ketoconazole inhibits the metabolism of cyclosporine in mice. In the present study we investigated the pharmacodynamics of the interaction and found that ketoconazole increased the immunosuppressive activity and toxicity of cyclosporine in vivo. Mice were treated orally with cyclosporine (25-200 mg/kg/day), ketoconazole (100 mg/kg/day), drug vehicle, or drug combination for 14 days. Immunosuppression was studied by assaying radiometrically the delayed hypersensitivity response in mice that were sensitized and challenged to keyhole limpet hemocyanin. The dose-response curve for cyclosporine was shifted to the left when cyclosporine and ketoconazole were coadministered compared with treatment with cyclosporine alone. The dose that produced 50% of the maximal immunosuppression (ED50) for cyclosporine alone was 89 mg/kg/day, whereas the ED50 for cyclosporine, when combined with ketoconazole, was 33 mg/kg/day. Ketoconazole alone was not immunosuppressive. Toxicity was measured by the length of survival during a ten-day period of treatment. Animals receiving both drugs had a significantly lower survival: the ten-day survival of animals treated with 200 mg/kg/day of cyclosporine was 100% in the absence of ketoconazole treatment and 62.5% with 100 mg/kg/day of ketoconazole. The results of our study have two important implications. First, the effect of ketoconazole on cyclosporine immunosuppression is quantitatively similar to its effect on cyclosporine kinetics. Because of this two- to three-fold increase in both unchanged cyclosporine blood levels and immunosuppression, our results suggest that unchanged cyclosporine--not its metabolites--is the primary pharmacological agent of cyclosporine in vivo immunosuppression. Second, to avoid cyclosporine-induced nephrotoxicity, it has been recommended that the dosage of cyclosporine be decreased when this drug combination is used clinically. The results of this study suggest that a decrease in cyclosporine dosage will not lower the level of immunosuppression in these patients.