Studies have shown that the number of high-frequency QRS notches increases after myocardial infarction (MI). To assess overall high-frequency (greater than 80 Hz) potentials more quantitatively, we adapted filtered and the root-mean-square (RMS) voltage of the residual (80-300 Hz) signal computed. High-frequency RMS values were significantly (p less than 0.01) greater in leads II, III and aVf in normal subjects (n = 12) than in patients with inferior infarction (n = 12). Similarly, high-frequency RMS values were higher (p less than 0.01) in leads V2 and V5 in normal subjects (n = 14) than in patients with prior anterior MI (n = 14). A reduction in high-frequency RMS values with inferior infarction was independently confirmed using Fourier analysis of the QRS in lead II. QRS notching in these subjects was also quantified by computing the number of baseline crossings of the first derivative (dV/dt). As predicted, notching was significantly greater (p less than 0.05) both with inferior MI (lead II) and anterior MI (lead V5). However, contrary to classic theory, the number of notches correlated negatively with direct measurements of high-frequency RMS voltage in lead II (r = -0.63) and lead V5 (r = -0.49). Positive correlations were obtained between high-frequency potentials and two new indexes that measure the amplitude of QRS dV/dt-peak-to-peak amplitude of dV/dt and RMS dV/dt. Using these indexes, absolute separation of inferior MI patients and normal subjects was obtained. We conclude that MI increases low-amplitude QRS notching but diminishes total high-frequency voltage, probably because of an overall decrease in electromotive potentials and slowing of ventricular conduction.