In 1091 patients with isolated Björk-Shiley, Lillehei-Kaster, Starr-Edwards, and St. Jude Medical mitral and aortic valve replacement, hemolysis parameters were determined (hemoglobin, LDH, haptoglobin, free plasma hemoglobin, reticulocyte count, serum bilirubin, transferrin, urine hemosiderin, schistocyte count). In 1006 of these patients no valve dysfunction was detected, while in 85 patients either paravalvular leakage or a thrombosis of the prosthetic valve was present. Haptoglobin was the most sensitive parameter for detecting even mild intravascular hemolysis, which was present in two-thirds of patients after alloprosthetic heart valve replacement. For quantifying red cell damage LDH was useful. Hemolysis was somewhat higher after Lillehei-Kaster and Starr-Edwards than after Björk-Shiley or St. Jude Medical implantation. The variance of LDH levels can be explained in a high percentage by correlating them with the hemodynamic findings at rest and exercise, which are indirect parameters of velocity profiles. Hemolysis is higher after aortic than after mitral valve replacement, with the exception of St. Jude valves. In patients with perivalvular leakage or valve thrombosis, red cell damage is more pronounced than in normally functioning prostheses (p less than 0.0005). When the hemolysis characteristics of the individual types of prosthesis are taken into account, the degree of hemolysis is a reliable indicator (p less than 0.05) of the functional integrity of the prosthesis. However, the degree of hemolysis does not correlate with the hemodynamic significance of perivalvular regurgitation.