Mitral annuloplasties [255] using 43 Carpentier rings and 212 flexible rings were performed. Simultaneously, 307 patients with predominant insufficiency were treated with a Hancock bioprosthesis. A critical analysis of the flexible ring group vs. the Hancock group is presented. The average ages were 41.3 vs. 43.3 yr. Preoperatively, 55 vs. 60% were in functional class III and 9.6 vs. 25.4% were in class IV. The valve lesion was caused by rheumatic disease in 91.2 vs. 93.9%. A multiple-valve operation was required by 51 vs. 60%. There was a significant difference in the operative mortality rates (1.8 vs. 11.4%). This difference was maintained for each preoperative functional class and for the isolated and multiple valve groups. There was no significant difference in the late mortality rates (1.4 vs. 3.7%). The total actuarial survival rates were 96.4 and 81% (maximum follow-up times 4.0 and 4.5 yr). The incidence of embolism was 2.4% and 4.1% per patient-year, with both groups being subjected to the same anticoagulation policy. The incidence of dysfunctions was equal (3.9% per patient-year). Postoperative hemodynamic evaluation of 72 patients having annuloplasty and 129 patients having Hancock valve replacement showed basal mean transmitral gradients of 10.98 .+-. 3.6 and 9.66 .+-. 2.73 mm Hg. The average effective orifice areas were 1.93 .+-. 0.74 and 2.25 .+-. 0.46 cm2. While the bioprosthesis behaves hemodynamically like a flow-related variable orifice, the reconstructed valve orifice is dependent upon its preoperative anatomy. The comparison of flexible ring annuloplasty and Hancock valve replacement for mitral valve disease shows a lower operative mortality for the former and comparable medium-term postoperative course. Since both techniques have a limited durability, longer follow-up periods will soon resolve this controversial issue.