The effects of desmopressin on postoperative bleeding and postoperative transfusion requirements were studied in ninety-two hemostatically normal patients who had had an elective primary total hip or total knee arthroplasty. The patients were randomized into either a placebo or a desmopressin group in a double-blind prospective clinical trial. During closure of the wound, desmopressin (0.03 microgram per kilogram of body mass) or the placebo was infused into a peripheral vein over a twenty-minute period. Compared with the placebo, desmopressin did not significantly decrease blood loss or transfusion requirements, and it did not affect the postoperative platelet or fibrinogen levels or the bleeding time. The results were no different even when the treatment and control groups were matched according to surgeon, use of cement for the femoral and knee components, preoperative use of non-steroidal anti-inflammatory agents, or performance of a lateral release for total knee arthroplasty. We concluded that desmopressin does not reduce blood loss or transfusion requirements after total joint arthroplasty.