Abstract OBJECTIVES This study provides an overview of the change over a 45-year time period in the characteristics and outcome of patients with tricuspid valve disease undergoing surgical tricuspid valve replacement (TVR). METHODS The characteristics and outcomes of all consecutive TVRs from November 1972 to November 2017 at Erasmus MC were collected retrospectively. A logistic regression analysis was conducted to identify the significant predictors of 30-day mortality. Multivariable Cox regression analysis was used to identify the potential risk factors of patient outcome and the effect of time on these factors. RESULTS Ninety-eight patients with tricuspid valve dysfunction underwent 114 consecutive TVRs at a mean age of 50.1 ± 17.2 years (68.5% female). Aetiology changed over time from predominantly functional regurgitation (42.9% in 1972-1985) to predominantly carcinoid heart disease (47.7% in 2001-2017). Early mortality declined significantly from 35% in 1972–1985 to 6.7% in 2001–2017 (P < 0.001). Over time, the hazard ratio of late mortality decreased for higher New York Heart Association class, lower preoperative haemoglobin, and high central venous pressure and increased for the presence of preoperative leg oedema, higher creatinine and alkaline phosphatase. The late survival was 43.8% ± 5.89% at 10 years and was comparable among eras (P = 0.44). The cumulative incidence of reoperation at 10 years was 14.1% (2.3–26.0) in biological valves and 4.9% (0.1–10.3) in mechanical valves (P = 0.25). CONCLUSIONS Patient characteristics, potential risk factors and patient outcome changed considerably over time in patients undergoing TVR. Notably, there was a shift in aetiology, completely altering the patient population and their characteristics.