Left ventricular (LV) mass is determined to a large extent by body size. This has created controversy regarding the optimal method of defining normal values in the clinical setting. Previous groups have advocated indexing LV mass for body surface area, lean body mass, or height, which is an obesity-independent measure of body size. This study describes a new approach that involves dividing LV mass by height raised to a noninteger power. In a sample of 387 men and 714 women who were between 20 and 45 years of age and who were free of cardiovascular disease, the height exponent was determined by logarithmic regression models to be 2.12 in men and 1.91 in women; in a pooled analysis, the height exponent was 1.97. This approach reduced male versus female differences in the mean and 95th percentile values for LV mass; mean value differences who reduced from 52% among raw values to 29% among adjusted values. Compared with height or body surface area indexation, it was found to slightly increase the correlation between LV mass and systolic blood pressure. In contrast, body surface area indexation reduced the association between LV mass and obesity. The current method of indexation is independent of obesity, reduces LV mass variability associated with body size and gender, and may therefore be a useful method for defining normal and abnormal values of LV mass in the clinical setting. The findings of this study also suggest that indexation of LV mass for body surface area is inappropriate.