Valve-related complications after Björk-Shiley mitral (n = 475), aortic (n = 424), or mitral-aortic implantation (n = 119) were compared to complications after St. Jude mitral (n = 173), aortic (n = 152), and St. Jude mitral and aortic (n = 63) replacements. The 1,018 consecutive patients with Björk-Shiley valves had been operated upon between 1974 and 1982, those with St. Jude valves between 1978 and 1982. All patients were placed on anticoagulant therapy with phenprocoumon early after operation and no significant intergroup differences in the effectiveness of the anticoagulant therapy were found. At a comparable follow-up time of approximately 23 months, 24 major thromboembolic episodes were observed after Björk-Shiley mitral (BSM) and 3 after St. Jude mitral valve implantation (SJM), corresponding to a thromboembolic rate of 2.82/100 patient years with BSM and 0.93/100 patient years with SJM. After aortic valve replacements, 1.93 events in 100 patient years occurred after Björk-Shiley aortic (BSA) and 0.73 after St. Jude aortic implantation (SJA). In patients with double valve replacements, these rates were 3.2 (BSM + BSA) and 0.88 (SJM + SJA), respectively. The cerebral vessels were involved in 52% and the arteries of the extremities in 22% of these major events. Six Björk-Shiley prostheses had to be replaced because of valve thrombosis. The overall incidence of severe hemorrhagic complications was 2.94/100 patient years in BSM and 1.79 in SJM. After aortic valve replacement, we found rates of 1.80/100 patient years (BSA) and 2.57/100 patient years (SJA), respectively. Intravascular hemolysis no longer seems to be a significant clinical problem. However, indications of red cell damage after heart valve replacement were significantly greater in patients with perivalvular leakage, valve thrombosis, or dysfunction than in those with normally functioning prostheses. Reoperations were necessary because of valve thrombosis (0.46%), perivalvular leakage (2.2%), or prosthetic valve endocarditis with concomitant perivalvular regurgitation (0.46%). One valve had to be replaced because of fracture of the outlet strut of a BSM prosthesis. Hemorrhage due to the anticoagulant treatment was thus the most frequent complication, without significant intergroup differences, while thromboembolic complications were significantly more frequent after Björk-Shiley mitral, aortic, and double valve replacements than after St. Jude implantation. This may lead to the consideration of a change in the prophylaxis of thrombus formations with the St. Jude valve, especially in aortic valve replacements with sinus rhythm.