Lipoprotein profiles were measured before and two months after complete withdrawal of prednisone in 34 kidney and 9 kidney-pancreas transplant recipients subsequently maintained on cyclosporine and azathioprine. Withdrawal of steroid therapy was accompanied by a 17% reduction in total serum cholesterol levels. However, there was a parallel reduction in all other measured lipoprotein concentrations, including an 18% reduction in high-density lipoprotein cholesterol levels. In diabetic recipients of a kidney or kidney-pancreas transplant, the ratio of total to high-density lipoprotein cholesterol was unchanged after steroid withdrawal. In nondiabetic kidney transplant recipients, this ratio actually increased significantly following withdrawal of prednisone. These observations suggest that it is premature to presume that withdrawal of steroid therapy will reduce the cardiovascular risk related to hyperlipidemia in cyclosporine-treated kidney or kidney-pancreas transplant recipients.