Continued adverse remodeling of myocardium after infarction may lead to progressive ventricular dilation and heart failure. We tested the hypothesis that exercise training in a healed myocardial infarction-dysfunction rat model can favorably modify the adverse effects of ventricular remodeling including attenuation of abnormal myosin gene expression. Sprague-Dawley rats were subjected to either proximal LAD ligation or sham operation. At 5 wk after the operation, animals were randomly assigned to sedentary conditions or 6 wk of graduated swim training, creating four experimental groups: infarct sedentary (IS), infarct exercise (IE), sham sedentary (SS), and sham exercise (SE). At 11 wk all rats were sacrificed and analyzed. Compared to sedentary infarct controls, exercise training attenuated left ventricular (LV) dilation and allowed more hypertrophy of the non infarct wall. The exercise-trained hearts also showed a reduction in the estimated peak wall tension. Northern blot analysis showed an increase in beta-myosin heavy chain expression in the hearts of the sedentary infarction group soon after infarction when compared to sham controls. However, with exercise training, there was a significant attenuation of the beta-myosin heavy chain expression in the myocardium. Exercise training in a model of left ventricular dysfunction after healed myocardial infarction can improve the adverse remodeling process by attenuating ventricular dilation and reducing wall tension. The abnormal beta-myosin expression was also attenuated in the exercise trained group. This is evidence that abnormal gene expression following severe myocardial infarction dysfunction can be favorably modified by an intervention.