To study the effectiveness and nephrotoxic side-effects of cyclosporin A (CsA) in renal transplant recipients, a prospective randomised trial was designed to compare CsA with azathioprine (Aza). Each treatment group consisted of 40 patients; in the CsA group, 18 were randomly selected for conversion to Aza after 3 months. The 1-year graft survival for CsA-treated patients was 87% compared with 66% for the Aza group (P = 0.033). Anti-rejection therapy was administrated to 78% of the patients in the Aza group and 47% of those in the CsA group (P less than 0.01). There was no difference in the incidence of primary non-functioning kidneys, cytomegalovirus infections, hypertension, or degree of proteinuria between the two treatment groups. At 3 months the mean creatinine clearance was 42 +/- 2 ml/min (mean +/- SEM) for the CsA group compared with 56 +/- 4 ml/min for the Aza group (P less than 0.01), whereas the mean creatinine clearances at 6 months for both the converted and the non-converted CsA-treated patients did not differ from that found in the Aza-treated group. At 1 year, the mean creatinine clearance for CsA-treated patients who were converted to Aza was higher than that found for Aza-treated patients (62 +/- 7 vs 50 +/- 6 ml/min; P less than 0.05). Furthermore, the increment in creatinine clearance observed after conversion from CsA to Aza at 3 months showed a linear relationship (r = 0.9061) with the CsA trough levels before discontinuation of the drug. This indicates that CsA treatment induces a dose-dependent, nephrotoxic side-effect which is probably reversible.