Percutaneous transvenous mitral commissurotomy was performed in 106 consecutive patients. Significant symptomatic improvement was achieved in 97 patients (92%). Mean left atrial pressure decreased (from 18 +/- 8 to 11 +/- 8 mm Hg, p less than 0.00001), mean mitral diastolic pressure gradient decreased (from 12 +/- 7 to 7 +/- 6 mm Hg, p less than 0.00001), and mitral valve area increased (from 1.40 +/- 0.40 to 2.00 +/- 0.50 cm2, p less than 0.00001). Based on echocardiographic characteristics of the mitral apparatus, patients were grouped retrospectively in three categories: pliable (group 1, n = 37), semipliable (group 2, n = 59), and rigid (group 3, n = 10). Clinical success was achieved in 36 patients of group 1 (97%) and in 55 patients of group 2 (93%). Only six patients in group 3 (60%) improved symptomatically (p less than 0.001 vs. group 1, p less than 0.001 vs. group 2). The severity of mitral regurgitation increased in five patients of group 1 (14%), in 12 of group 2 (20%), and in three of group 3 (33%). Six patients had recurrent symptoms at 9 months after commissurotomy. Recurrence of symptoms was significantly more frequent in group 3 compared with the other two groups (group 1, 3%; group 2, 4%; and group 3, 50%; p less than 0.0001 vs. groups 1 and 2). Multiple regression analysis identified the previously mentioned echocardiographic characteristics of the mitral apparatus as the significant predictor for clinical outcome. Thus, percutaneous transvenous mitral commissurotomy can be considered a safe and effective treatment for patients with pliable valves. Patients with semipliable or with rigid valves should be selected for operation very carefully.