The incidence and severity of infectious complications were retrospectively investigated in 80 renal transplant patients who had been selected randomly to receive either cyclosporine (CsA) or azathioprine (Aza) in combination with low doses of corticosteroids. In the first 3 months, the incidence of infections was twice as high in the Aza-treated patients (p less than 0.05) which was due to an increase of predominantly minor infections. This increased incidence of infections was related to the increased number of anti-rejection treatments in the Aza-treated group. The types of infections were not different between the 2 treatment groups nor did the incidence of CMV infections differ. In the CsA-treated group, patients were randomly assigned at 3 months after transplantation to either continuation of CsA therapy or conversion to Aza. After conversion a small but not significant increase in predominantly minor infections was observed, which may be attributable to increased doses of corticosteroids given during the conversion. We conclude that in the first 3 months following transplantation, CsA therapy is associated with significantly less infections than Aza therapy; following conversion of CsA to Aza at 3 months only a small increase in the infection incidence is found.