This study compared immediate with long-term results of mitral balloon valvotomy (MBV) in patients who underwent MBV as an initial procedure versus those who underwent repeat MBV. Fifty-six patients who were a mean age of 28 +/- 8.8 years (group A) and had mitral restenosis after surgical or balloon commissurotomy underwent MBV and were compared with 524 patients who were a mean age of 31 +/- 11 years (group B) and underwent MBV as an initial procedure. Prospective data obtained included demographic, hemodynamic, echocardiographic, and clinical follow-up for 0.5 to 15 years after MBV. No deaths or technical failure were encountered after MBV. Mitral regurgitation >2/4 occurred in 1 patient (2%) in group A and in 9 patients (2%) in group B (p = 0.24). Fifty-two of 56 patients (93%) in group A had good immediate results (mitral valve area >or=1.5 cm(2), mitral regurgitation <2/4), as did 504 of 524 patients (96%) in group B (p = 0.4). Actuarial values for freedom from restenosis at 10 years were 58 +/- 7% for group A versus 69 +/- 3% for group B (p = 0.18). Ten-year event-free survival rate was 54 +/- 7% for group A versus 80 +/- 3% for group B (p <0.005). The predictors of event-free survival were age (p = 0.003), echocardiographic score (p <0.0001), and baseline atrial fibrillation (p = 0.01). In conclusion, MBV is safe and provides good immediate results in patients who have restenosis. Long-term results are inferior compared with de novo mitral stenosis but is still satisfactory. More than 50% of patients remained improved at 10 years, thus enabling the operation or reoperation to be deferred.