OBJECTIVE To evaluate the utility of cyclosporine (CsA) trough concentrations as a monitoring tool for acute graft rejections and CsA nephrotoxicity. METHODS Retrospective chart review. METHODS University-affiliated teaching hospital. METHODS One hundred thirty-seven adults who had undergone kidney transplantation. RESULTS Clinical data extracted from the charts were CsA dosage, CsA trough levels (whole blood, HPLC method), biopsy findings to confirm acute rejections, and serum creatine to determine clearance by the Jelliffe method. Data were collected at up to 1 month, between 1 month and 3 months, and between 3 and 12 months after transplantation. For each time period, receiver's operating characteristics curves were generated to identify the optimum CsA concentration for avoiding acute rejection and CsA nephrotoxicity. At up to 1 month, the CsA therapeutic response threshold was 182 ng/ml (sensitivity 69%, specificity 84%, p<0.0001) and toxicity threshold for CsA nephrotoxicity was 204 ng/ml (sensitivity 89%, specificity 56%, p<0.0001). Between 1 month and 3 months, the respective figures were 175 ng/ml (sensitivity 58%, specificity 89%, p<0.0002) and 189 ng/ml (sensitivity 87%, specificity 65%, p<0.0001). Between 3 and 12 months, the CsA therapeutic response threshold decreased to 135 ng/ml (sensitivity 56%, specificity 40%, p>0.1) and the toxicity threshold for CsA nephrotoxicity remained relatively static at 204 ng/ml (sensitivity 100%, specificity 14%, p<0.0001). CONCLUSIONS Early in CsA therapy it is essential to prevent graft rejection. Drug concentrations exceeding approximately 182 ng/ml threshold accomplish this goal. Later, successful therapy demands that CsA nephrotoxicity be avoided. This goal is accomplished by not exceeding a CsA concentration of 204 ng/ml.