In 33 patients, including 12 control subjects and 21 with eccentric LVH, LV mass determined by angiocardiogram was correlated to 26 VCG measurements (Frank system) calculated from the scalar X, Y, and Z leads. The results demonstrated that the most reliable indices of VCG in assessing the severity of eccentric LVH determined by angiocardiogram were the magnitude of the spatial mean QRS vector and the time of the spatial maximal QRS vector ("spatial VAT"), of which correlation coefficients were 0.93 and 0.93, respectively. Such high correlation coefficients have never been obtained with the usual ECG analysis. These findings strongly suggest that (1) increased QRS voltage and usual prolonged QRS duration in eccentric LVH are due to an increase in LV mass, and (2) prolonged VAT observed in eccentric LVH is closely related to an anatomic alteration, namely, the greater distance of intra-ventricular conducting pathways as the result of LV dilatation, as an increase in LV mass is usually paralleled by the grade of the chamber enlargement in this type of LVH. Regarding the T loop, correlations between the LV mass and the VCG measurements were less as compared to those of the QRS loop. In general, T changes in moderate or severe LVH may be also related to a certain altered cardiac muscle state, in addition to an increase in LV mass. Angiocardiographic and light microscopic findings of a patient with eccentric LVH in whom a widened QRS-T angle was demonstrated to an extent much more than that expected with an increase in LV mass are presented and discussed. The spatial pattern analysis by VCG is very useful and reliable in assessing the severity of eccentric LVH.