Hemolysis was assessed in 86 patients after mitral valve replacement. Twenty-four patients had mitral valve replacement with a Björk-Shiley valve, 32 patients with a Starr-Edwards composite-seat valve, and 30 patients with an irradiated frame-mounted aortic homograft valve. Hemolysis was determined by red cell survival and autologous 51Cr-tagged red cells, LDH, serum haptoglobin, hemosiderinuria, reticulocyte count, red cell fragment count, and hemoglobin estimation. The degree of hemolysis was classified as mild, moderate, or severe. Thirty-five per cent (nine of 24) of the Björk-Shiley group showed mild hemolysis, whereas 85 per cent (28 of 32) of the Starr-Edwards group had evidence of hemolysis-of these, a third were mild and the rest moderate. The homograft series did not show any comparable evidence of hemolysis. Statistical analysis of the parameters of the study comparing homograft with Björk-Shiley valves showed no significant difference except in red cell survival, which showed a highly significant difference (p less than 0.001). Comparing homograft with Starr-Edwards valves, all parameters showed highly significant differences (p less than 0.001).