Percutaneous mitral valvotomy is an alternative to surgery for the treatment of selected patients with mitral stenosis. With the purpose of reviewing our own experience and to determine the relationship of several variables with procedural success or complications rate, we have analyzed the immediate results of 335 consecutive procedures (90% with the Inoue balloon). Mean age was 51 +/- 12 years, 79% were female, 59% were in atrial fibrillation, 49% in NYHA class > or = III and 34% had mild mitral regurgitation. Mean echocardiographic score was 7.6 +/- 1.9 and up to 25% of patients had a total score > or = 8. Mean mitral valve area (Gorlin) increased from 0.94 +/- 0.2 to 1.87 +/- 0.5 cm2, with increments > or = 50% in mitral area in 88% and a final area > or = 1.5 cm2 in 85% of patients. Patients with an score > 8 obtained smaller mitral areas (1.64 +/- 0.4 vs 1.95 +/- 0.5 cm2. p < 0.01), and a higher percent of suboptimal (< 1.5 cm2) mitral area (22 vs 2%, p < 0.001) than patients with a lower score. Echocardiographic score (p = 0.009), balloon size (p = 0.01) and left atrial diameter (p = 0.04) were identified as independent predictors of a post-procedural mitral area > or = 1.5 cm2. Complications including cardiac tamponade (3 cases), cerebral ischemic events (1 case) and death (1 case) were rare. Mitral regurgitation increased by 2 or more grades in 14%, but regurgitation was severe in only 6% and early mitral surgery was required in 6 (2%) patients. No clinical, anatomical or procedural variables were found to be predictors of a significant increase in mitral regurgitation, a complication which occurred even in valves with a low echocardiographic score. In conclusion, percutaneous mitral valvotomy is a safe and effective procedure for the treatment of selected patients with mitral stenosis. Since complications, even if rare might be serious, the procedure should be performed only in symptomatic patients.