The clinical performance of tricuspid valve replacement with bioprostheses and mechanical prostheses was assessed in a series of 5,489 total valve replacement operations performed from 1975 to 1992. There were 97 (1.8%) tricuspid valve replacements in 94 patients (16 men, 78 women) with a mean age of 55.4 +/- 13.8 years. Bioprostheses (mean patient age, 55.9 +/- 14.1 years) were used in 83 operations and mechanical prostheses (mean patient age, 52.1 +/- 11.9 years) were used in 14 operations. There were 30 isolated tricuspid valve replacements and 67 tricuspid valve replacements incorporated in multiple valve replacements. The total cumulative follow-up was 360 patient-years (bioprostheses, 321 years; mechanical prostheses, 39 years) (96.8% complete). The mean follow-up was 3.7 years (bioprostheses, 3.9 years; mechanical prostheses, 2.8 years) (p = not significant). The early mortality was 14.4% (bioprostheses, 14.5%; mechanical prostheses, 14.3%) (p = not significant) (isolated replacement, 13.3%; multiple replacement, 14.9%). The late mortality was 9.2% per patient-year (isolated replacement, 12.2% per patient-year; multiple replacement, 7.9% per patient-year). The freedom from structural valve deterioration at 5 and 7 years was 100% for mechanical prostheses and 97.1% +/- 2.9% for bioprostheses (p = not significant). For isolated tricuspid valve replacement, the freedom from structural valve deterioration for bioprostheses was 90.9% +/- 8.7% at 5 years and at 7 years (p = not significant). For the mechanical prostheses, the freedom was 100%.(ABSTRACT TRUNCATED AT 250 WORDS)