Bleeding is a major complication of uremia. Both cryoprecipitate and desmopressin effectively shorten the prolonged bleeding time and favorably influence clinical bleeding, but the former carries the risk of transmitting blood-borne infectious diseases, and both cryoprecipitate and desmopressin have a short duration of action. Preliminary evidence has suggested that estrogens may be useful, and we therefore performed a randomized, double-blind, crossover trial comparing the effect of conjugated estrogens with that of placebo on hemorrhagic tendencies and the bleeding time in six patients with uremia who were on maintenance hemodialysis. Five daily infusions of placebo or conjugated estrogens were administered at the beginning of one-month trial periods. Estrogen shortened the bleeding time in all six patients. The effect was detectable six hours after the first infusion, reached its maximum in all patients between days 5 and 7, and lasted for 14 days. By day 16 after the last infusion, the bleeding time had returned to base line in four of the six patients. No side effects were noted during or after estrogen infusion. Estrogens did not influence the circulating level of von Willebrand factor or change its multimeric structure. Moreover, the defective platelet aggregation and thromboxane formation observed in the patients were not corrected by estrogens. We conclude that conjugated estrogens are an adequate alternative to cryoprecipitate or desmopressin for the treatment of bleeding associated with renal failure, especially when a longer duration of action is needed and immediate onset of the effect is not essential. The mechanism of action of estrogens remains to be clarified.