A clinical trial with the Medtronic-Hall (M-H) valve was started in November 1981. From then until December 1982, 157 patients underwent heart valve replacement: Mitral valve replacement (MVR) 63 patients, isolated aortic valve replacement (MVR) 41 patients, and 53 underwent combined mitral-aortic valve replacement (MVR + AVR). Hospital mortality was 2.5% overall (4/157). The total follow-up was 147 patient-years (range 1 to 12 months). Late mortality was 3.2% for MVR, 2.5% for AVR and 1.9% for double MVR + AVR. Oral anticoagulation therapy (Dicumarol) was administered to all patients included in this study. There were no anticoagulant complications and no prosthetic bacterial endocarditis registered in this series of patients up to date. The incidence of systemic embolism was 2.5 events per 100 patient-years in the AVR group. No embolic episodes have been registered in the other 2 groups. The incidence of valvular dysfunction (expressed as events per 100 patient-years) were: --Perivalvular leak (0 MVR, 0 AVR and 2 MVR + AVR) --Valve thrombosis (1.7 MVR, 0 AVR and 0 MVR + AVR) --Mechanical dysfunction (1.7 MVR, 0 AVR and 0 MVR + AVR) Despite a short follow-up, the results with the Medtronic-Hall valve are encouraging and compare favorably with other mechanical prostheses.