Administration of anticonvulsant medication to recipients of cadaver renal allografts appears to be associated with decreased allograft survival. The one-and two-year actuarial graft survival rates for 20 index grafts was significantly lower than for 92 control grafts. Since phenobarbital and diphenylhydantoin increase the metabolism of corticosteroids, it is proposed that renal allograft recipients receiving these anticonvulsants may have ineffective immunosuppression, leading to a higher incidence of graft failure.