The long-term results of conversion of cyclosporine to azathioprine and those of continuous CsA therapy were evaluated in a prospective study of 66 renal transplant patients who had been randomly assigned to each treatment group at 3 months following transplantation. The start point of the study was thus at 3 months posttransplant; no differences in the three-year patient and graft survival were found; these consisted of 97% and 94% in the converted group and 100% and 94% in the nonconverted group, respectively. The incidence of one or more antirejection treatments did not differ between the two groups at 3-12 months (16% vs. 17%) or after 12 months (12% vs. 9%). The incidence of hypertension at different intervals ranged from 79% to 100% in the group on continuous CsA therapy versus 50 to 58% in the converted patients. The degree of proteinuria in the 2 groups was not different at at 12 months. At 24 and 36 months the proteinuria (g/24 hr) was higher in the converted group (0.51 +/- 0.18 and 0.53 +/- 0.13; mean +/- SEM) versus the CsA group (0.15 +/- 0.04 and 0.21 +/- 0.09). At 3 years, the mean creatinine clearance for the patients converted to Aza was higher than that found for the continuously CsA-treated patients (67 +/- 8 and 59 +/- 6 ml/min; mean +/- SEM). This study shows that early CsA conversion to Aza gives a slightly better 3-year graft function, although not significantly different, compared with continuous CsA therapy without differences in patient or graft survival.