OBJECTIVE Tricuspid valve replacement (TVR) has a high postoperative mortality, despite recent advances in perioperative care. We report the results of our experience in TVR with an emphasis on early mortality and morbidity and long-term follow-up. METHODS Between October 1994 and August 2007, 80 consecutive TVRs were performed in 78 patients. The mean age was 48+/-14 (range: 20-70) years. The underlying disease of the patients was classified as rheumatic (n=54), congenital (n=12), endocarditis (n=10) or degenerative (n=4). Previous cardiac surgery had been performed in 40 patients (50%). Isolated TVR was performed in 24 patients (30%). RESULTS Hospital mortality occurred in one patient (1.4%). Postoperative morbidities included intra-aortic balloon pump (n=5), bleeding re-operation (n=4), delayed sternal closure (n=3), acute renal failure (n=3), subdural haematoma (n=3), extracorporeal membrane oxygenation (n=1), mediastinitis (n=1) and pacemaker insertion (n=4). In 42 patients, ventilator support was needed for more than 72 h. Based on multivariate analysis, age (p<0.001) and the cardiopulmonary time (p=0.004) were the identified risk factors. Follow-up was completed in all patients with a mean duration of 56+/-37 (range: 0-158) months. During the follow-up period, there were seven deaths (8.8%), including five cardiac deaths. The 5- and 8-year survival rates were 95+/-3% and 79+/-9% and event-free survival rates were 76+/-6% and 61+/-9%, respectively. Based on multivariate analysis, the only identified predictors of late deaths was a postoperative low cardiac output (p=0.024). CONCLUSIONS TVR can be performed and low operative mortality can be achieved thorough optimal perioperative management in the current era.