The long-term effects of four different immunosuppressive regimens on renal allografts have been investigated up to four years. A total of 128 recipients of first cadaveric renal allograft were randomized, after an initial induction period, to receive either triple drug therapy--i.e., (low-dose) cyclosporine, azathioprine, and methylprednisolone, or any possible combination of two drugs--i.e., Aza plus CsA, Aza plus MP, and CsA plus MP. The actual four-year graft survival rates for the triple therapy, Aza plus CsA, Aza plus MP, and CsA plus MP groups were 72%, 69%, 75%, and 59%, and patient survival rates were 78%, 81%, 81%, and 84%, respectively, with no significant differences. The cumulative number of chronic rejections up to 4 years was 0.09, 0.29, 0.25, and 0.34 per patient per group (P = ns), respectively. At 2, 3, and 4 years posttransplantation, the graft function was significantly worse in the Aza plus MP group compared with the triple therapy group (P < .05). Of the 98 patients who did not have type I or II diabetes at the time of transplantation, 17 developed posttransplantation diabetes mellitus or an abnormal glucose tolerance test within 2 years posttransplantation. All these patients had received steroids and none of the patients without steroids had these abnormalities. At two years the mean cholesterol level was highest in the Aza plus MP group, 6.8 mmol/L and lowest in the group receiving triple therapy, 5.8 mmol/L (P = ns). The use of (low-dose) CsA was not associated with lipid abnormalities or with disturbances in glucose metabolism. A protocol graft biopsy was performed at two years on all functioning kidneys, and the histological changes were scored blindly. No CsA-specific changes, except isometric vacuolation in tubuli, were found. Histological alterations characteristic of chronic rejection were expressed as "chronic allograft damage index." Chronic allograft damage index was lowest in the triple therapy group, 1.5, compared with the other groups, 3.2-4.3 (P = .01), indicating the least histopathological change in the triple therapy group. In conclusion, this long-term study did not show any serious cyclosporine-related side-effects when used in low dose in combination with two other drugs. Some side-effects, such as posttransplant diabetes mellitus and probably some lipid abnormalities, could rather be traced to a higher dose of steroids. Moreover, the triple drug therapy was more efficacious than any double drug regimen in the prevention of chronic histological changes in renal allografts.