The urinary excretion of four enzymes (fructose-1,6-bisphosphatase, glutathione S-transferase, N-acetyl-beta-D-glucosaminidase and pyruvate kinase) was assayed daily in 59 patients following renal cadaveric allografting. 51 patients were given cyclosporin A (CyA group) as an immunosuppressive, 8 patients were treated conventionally with azathioprine and prednisolone (CON-group). Urinary enzyme output was evaluated by two different mathematical models. Model A follows single enzyme excretion, whereas model B also analyzes enzyme patterns. The best results were obtained by a combined analysis of all four enzymes with model B. In the CON-group the sensitivity was 1.00, the specificity 0.85, the predictive values of positive test 0.45 and all 12 graft rejections were diagnosed correctly. In the CyA group the sensitivity was 0.40, the specificity 0.99, the predictive value of positive test 0.33, and 6 out of 9 rejections were recognized. The evaluation of the single enzymes did not produce similarly good results with either model.