We performed radionuclide ventriculography before and within 1 week after percutaneous mirtal valvuloplasty (PMV) to evaluate left ventricular (LV) function in 20 patients(3 males and 17 females, mean age of 38 $pm$ 10 years) who were pure mitral stenosis before PMV and less than grade 1 mitral regurgitation developed after PMV. 9 out of 20 patients had atrial fibrillation and 3 patients developed a small left-to-right shunt (Ｑp/Qs<1.5) after PMV. PMV using double-balloon technique resulted in a increase in mitral valve area (0.9 $pm$ 0.3 to 2.1 $pm$ 0.8$ extrm{cm}^2$, p<0.001) and in rest cardiac output (4.2 $pm$ 1.0 to 4.8 $pm$ 1.4 L/min, p<0.005). And also a decrease in mean mitral gradient (16.2 $pm$ 7.0 to 5.2 $pm$ 3.0 mmHg, p<0.001) was noted. Comparisons of LV function by supine bicycle exercise radionuclide ventriculography before and after PMV showed no significant changes in rest LV ejection fraction (LVEF, 55.2 $pm$ 9.7 to 56.1 $pm$ 11.0%, p>0.05), maximal exercise LVEF (60.9 $pm$ 10.3 to 59.3 $pm$ 11.1%, p>0.05), peak ejection rate (2.02 $pm$ 0.58 to 2.15 $pm$ 0.60 EDV/ses, p>0.05), and peak filling rate (1.04 $pm$ 0.55 to 1.52 $pm$ 0.49 EDV/ses, p>0.05). However, LV end diastolic volume (LVEDV, -29.3 $pm$ 19.0 to +3.3 $pm$ 32.9%, p<0.001), stroke volume (SV, -22.0 $pm$ 22.0 to + 11.2 $pm$ 40.0%, p<0.001), and cardiac output (CO, 64.6 $pm$ 54.0 to 100.4 $pm$ 69.7%, p<0.05) with exercise compared to resting values were significantly increased. We conclude that PMV resulted in a significant increase in LVEDV, SV, and CO with exercise, but decreased LV systolic performance and no improvement in diastolic filling.