Background Late prognosis after successful percutaneous mitral commissurotomy (PMC) is unclear. We compared late results of PMC using Inoue versus double-balloon techniques up to 25 years in a randomized trial. Methods Between 1989 and 1995, 302 patients (77 men, 41 +/- 11 years) with severe mitral stenosis were randomly assigned to undergo PMC using Inoue (n = 152; group I) or double-balloon technique (n = 150; group D). The end points were the composite events of death, mitral surgery, repeat PMC, or deterioration of New York Heart Association (NYHA) class >= 3. Results During median follow-up of 20.7 years (maximum, 25.6), clinical events occurred in 82 (53.9%) patients in group I (37 deaths, 44 mitral surgeries, 9 repeat PMCs, 3 NYHA class >= 3) and in 79 (52.7%) patients in group D (34 deaths, 51 mitral surgeries, 5 repeat PMCs, 4 NYHA class >= 3). Event-free survival rates at 24 years were not significantly different between group I and group D (40.8% and 42.6%, respectively; hazard ratio [HR], 0.88; 95% confidence interval [CI], 0.65-1.20; P =.423). On multivariate analysis, absence of post-PMC commissural mitral regurgitation (MR) (HR, 1.84; 95% CI, 1.28-2.63; P =.001) and immediate post-PMC mitral valve area (MVA) b1.8 cm(2) (HR, 1.53; 95% CI, 1.04-2.25; P =.031) were independently correlated with clinical events after successful PMC. Conclusions The Inoue and double-balloon methods showed similar good clinical outcomes up to 25 years, and the achievement of effective commissurotomy to develop post-PMC commissural MR or immediate post-PMC MVA = 1.8 cm(2) is important in optimizing the late results of PMC.