Joint registry and hospital data bases for 5,024 total hip and total knee arthroplasties done between 1986 and 1988 at the Mayo Clinic were used to study prophylactic measures and frequency of symptomatic deep venous thrombosis and pulmonary embolism. In virtually all patients, graduated compression stockings were used, with or without another type of prophylaxis. Only 44 of 3,115 patients who underwent hip arthroplasty (1.4%) and 32 of 1,909 patients who underwent knee arthroplasty (1.7%) had definite or probable deep venous thrombosis or pulmonary embolism. Death definitely or possibly attributable to pulmonary embolism occurred in 11 patients who underwent hip arthroplasty (0.35%) and 1 patient who underwent knee arthroplasty (0.05%). Although patients with a history of deep venous thrombosis or pulmonary embolism were more likely to receive warfarin than were patients without such a history, the relative risk of symptomatic deep venous thrombosis or pulmonary embolism in patients who underwent hip arthroplasty and received warfarin postoperatively was approximately half that in patients who received other types of prophylaxis. The risk of death from pulmonary embolism was similarly diminished in the group that received warfarin. The lower rates of these complications in the patients who received warfarin support the prophylactic use of this agent after total hip arthroplasty.