The influence of exercise training on left ventricular function at rest (R), at anaerobic threshold (AT), and during peak exercise (PE) was evaluated in 12 healthy untrained and 13 trained (T) subjects who underwent Doppler echocardiography at R and radionuclide ventriculography at R and during exercise. The end-diastolic volume and stroke volume were significantly higher in the T group than in the untrained group at R. The ejection fraction rose significantly from R to AT and from AT to PE (80.0 +/- 0.84 vs. 83.6 +/- 0.91%), but no significant difference was observed between groups. The peak diastolic filling rate rose significantly during exercise, with a further significant increase observed in the T group (AT, 6.38 +/- 0.40 vs. 5.01 +/- 0.16 end-diastolic counts/s; PE, 8.24 +/- 0.42 vs. 7.15 +/- 0.35 end-diastolic counts/s). The percent variation of minimal systolic counts fell significantly at AT and PE in relation to R. Our data demonstrate that exercise training produces a significant increase in peak diastolic filling rate but no change in systolic function during exercise and that metabolic acidosis caused by exercise does not limit systolic function.