Vectorcardiograms (VCG) recorded in 37 subjects with marked chronic exogenous obesity (average weight: 285 lbs) were compared before and after a significant weight reduction (average weight loss: 86 lbs). They were also compared with the VCGs of 293 age and sex matched controls with normal body weight. The SVEC III corrected orthogonal lead system was used, and out of several hundred vectorcardiographic measurements obtained by computer processing, 59 measurements representing various scalar, planar and spatial voltage and angular measurements were selected for study and comparisons. No significant differences were found between the measurements of obese subjects and those of the controls. Of 59 VCG measurements, 13 showed significant differences after weight reduction (paired t test) although they remained within the range of normal controls. There was a trend toward decrease of P and QRS amplitudes after weight reduction. There were no significant changes in the angular measurements. Although they are statistically significant, these changes in voltage are too small to be detected in clinical vectorcardiography. The possible decrease of a preexisting myocardial hypertrophy superimposed onto the changes in the anatomy of the thorax might explain the effects of weight reduction. In general, the reproducibility of VCG measurements obtained by the SVEC III system was greater than that reported in day-to-day operation with the Frank system in normal subjects.