The usefulness of factor analysis as a means of describing in both general and specific terms an equivalent cardiac generator has been explored. High-fidelity, high-frequency electrocardiograms and vectorcardiograms of 31 normal subjects have been analyzed. The information in the electrocardiogram was reduced by factor analysis to the minimum number of independent waveforms which could account for all of this information. Because the first three principal factors might or might not represent a simple dipole, the efficiency of these three principal factors to resynthesize the originally recorded electrocardiograms was compared with the efficiency of McFee vectorcardiographic leads to perform the same resynthesis. The first three principal factors contained more information than that contained in the vectorcardiographic leads, suggestingthat the factors contain more than simple dipolar information. This would in turn suggest that a more complex generator than the dipole must be employed to explain the surface potentials derived from the standard electrocardiogram. With only two principal factors the precordial leads were rather easily reconstructed but reconstruction of leads I and VFwas considerably more difficult. Because the principal factors represent the most commonly occurring waveforms in the group, the less perfect resynthesis of leads I and VFwith two factors alone may result from lack of detection of multipolar component in the limb leads which have less in common with the precordial leads than the precordial leads do with each other.