Background. Little is known of time-related outcome and comparative performance of biological and mechanical prostheses following tricuspid valve replacement (TVR). Methods. A retrospective UK Heart Valve Registry study (Tan 1, 1986 to Tune 30, 1997) identified 425 patients who underwent TVR. Two-hundred twenty-five (52.9%) received biological and 200 (47.1%) received mechanical valves. One-hundred sixty (38%), 158, and 76 had isolated, double, and triple valve replacements, respectively. The follow-up was 96% complete with a total of 1,585 patient-years. Results. Thirty-day mortality for TVR was 17.3% (73 deaths). One-, 5-, and 10-year survival rates were 72.2%, 59.9%, and 42.9%, respectively. Year of operation (p = 0.04), age (p = 0.04), and number of valves implanted (p = = 0.03) predicted overall mortality. Age (p < 0.001) and year of operation (p = 0.002) predicted overall survival. Thirty-day mortality for biological and mechanical prostheses was 18.8% and 15.6%, respectively. One; 5-, and 10-year survival rates were 70.5%, 61.5%, and 47.7% for biological and 74.0%, 57.9%, and 33.9% for mechanical prostheses, respectively. Freedom from reoperation at 1 and 10 years was 98.7% and 97.4%. Freedom from death or reoperation was 71.2% at 1 year and 41.9% at 10 years. None of the above outcomes was significantly different between the type of valve prostheses. Conclusions. TVR carries a high 30-day mortality and a poor longer term survival. No superiority could be identified for biological or mechanical prostheses in the tricuspid position for either survival or reoperation. (C) 1998 by The Society of Thoracic Surgeons.