Magnetic field maps and body surface potential maps can be used to measure cardiac activity. The ability of magnetic and potential body surface maps to identify patients' vulnerable to recurrent sustained ventricular arrhythmia (VA) were compared. Magnetic field maps (MFM) and body surface potential mapping (BSPM) were obtained from 76 normal (N) subjects, 15 myocardial infarct (MI) patients, and 15 VA patients. QRST integral maps were calculated for each subject and nondipolar content was determined using Karhunen-Loeve transform eigen-maps. Although differences in nondipolar content were significant between the normal and patient groups (P = 2.4 x 10(-5) for BSPM and P = 6.0 x 10(-8) for MFM), differences in nondipolar content between MI and VA patients using QRST integral BSPM and MFM maps were not significant. The trajectory of the location of the maxima and minima on the map area during the QRS and ST-T intervals were also constructed. Discrimination between MI and VA patients was based on intergroup differences in the amount of fragmentation of the trajectory plots. The ST-T trajectory plots were significantly more fragmented (P < 0.0001 and P < 0.05 for MFM and BSPM, respectively) for VA than for MI patients. The ST-T interval MFM and BSPM trajectory plots enabled separation of MI and VA patients with accuracies of 83% and 73%, respectively. These results suggest that repolarization MFM and BSPM extrema trajectory plots can be used effectively as a means of identifying patients at risk for VA.