In the period between January 1979 and January 1989, 72 patients (46 women and 26 men, average age 46 years) out of 697 patients undergoing a tricuspid valve procedure. underwent prosthetic valve replacement for organic tricuspid valve disease. Sixteen patients had pure tricuspid regurgitation (Group 1). The causal pathology was essentially traumatic (6 cases) and infections (endocarditis) (5 cases). Fifty-six patients (Group II) had organic tricuspid valve disease associated with a left-heart valvular lesion. The causal pathology was rheumatic fever in 52 cases and endocarditis in 4 cases. In 40 patients (71.5 %) the procedure was a reoperation of a mitral and/or aortic and/or tricuspid valve prosthesis. In 21 cases, the initial operation had involved the tricuspid valve. All patients in Group I underwent valvular replacement with a bioprosthesis. In Group II, 40 patients were implanted with a bioprosthesis and 16 with a mechanical valve prosthesis. The follow-up was complete for an average period of 7 years (range 2 to 12 years) for a total of 295 patient-year. The hospital mortality was nil in patients with isolated tricuspid valve disease. Reoperation for deterioration of a porcine bioprosthesis was required after 4 years in 1 case. Terminal cardiac failure was responsible for one late death 7 years after surgery. The hospital mortality was 26.7 % (15 patients) in Group 11. The actuarial rate of patients free of reoperation at 7 years was 80 +/- 8 %. The actuarial rate of patients free of any prosthetic valve related complication was 78 +/- 2 %. The 7 year survival rate excluding operative mortality was 65 +/- 8 %. These results show that when no conservatice procedure is feasable, tricuspid valve replacement can be performed with a low rate of valve-related complications. The poor prognosis of this surgery is mainly due to the presence of left heart disease. Long-term follow-up is required to assess the results of tricuspid valve replacement with bioprostheses.