To determine whether the departure parameters derived from a "departure loop" of a vectorcardiogram are more accurate than conventional spatial parameters in evaluating myocardial infarct size, 74 patients with first-onset myocardial infarction (MI) were studied. The correlation between the departure parameters (amplitudes in scalar leads of the departure loop) and the percent defect volume of thallium myocardial scintigrams (%DV) was compared with that of the spatial parameters (magnitude, azimuth, and elevation of the original QRS loop). In anteroseptal MI, the amplitude of a 20-msec vector in the z-axis and the azimuth of a 30-msec vector (H30) were significantly correlated with %DV (r=0.783, p<0.001 and r=0.572, p<0.05). In anteroseptal MI with involvement of the lateral wall, the amplitude of a 30-msec vector in the x-axis and H30 showed significant correlation with %DV (r=0.802, p<0.001 and r=0.772, p<0.01). In inferior and inferoposterior MI, the amplitude of a 30-msec vector in the y-axis and the elevation of a 30-msec vector were significantly correlated with %DV (r=0.920, 0.891, p<0.001 and r=0.871, 0.678, p<0.01, respectively). In conclusion, the departure parameters are more accurate than the spatial parameters for evaluation of myocardial infarct size.