Prothesis choice in isolated mitral valve replacement for patients aged 75 years or younger remains debated as most studies comparing prothesis type have included large proportions of combined operations and benefits are influenced by concomitant procedures. This study compared long-term outcomes of isolated mechanical versus bioprosthetic mitral valves in different age groups of propensity-matched populations. This is a retrospective, multicentre, propensity-matched observational study. Baseline characteristics, operative details and long-term outcomes (mortality and freedom from surgical/transcatheter reintervention) were collected. Totally, 1536 isolated mitral valve replacements (806 mechanical, 730 bioprosthetic) were performed between 2000 and 2017. Over 90% of eligible patients successfully underwent propensity matching, yielding 226 each of mechanical and bioprosthetic valves in patients aged <65 years and 171 each of bioprosthetic and mechanical valves in patients aged 65-75 years with median follow-up of 13 years (maximum 20 years). In matched patients <65 years, 10-year survival was superior with mechanical valves versus bioprosthetic valves (78.2% vs 69.8%, P = 0.029), as was 10-year freedom from reintervention (96.2% vs 81.3%, P < 0.001). For matched patients between 65 and 75 years, there were no differences between mechanical and bioprosthetic valves in 10-year survival (64.6% vs 60.8%, P = 0.86) or 10-year freedom from reintervention (94.0% vs 97.2%, P = 0.23). Rates of post-operative stroke, gastrointestinal bleeding, renal failure and permanent pacemaker insertion were similar. In patients requiring isolated mitral valve replacement, mechanical valves confer significantly better long-term survival and freedom from reintervention for patients <65 years, while no benefit is observed at age 65-75 years compared to bioprosthetic valves.