The quantitative relationships that exist between left ventricular (LV) wall shortening, wall thickening, and geometry during LV ejection are not well defined. We used a mathematical model to measure these parameters in 40 patients with various LV geometries studied by echocardiography. As opposed to wall shortening, the percent contribution of wall thickening to LV ejection (% DELTA-V(h) was 25 +/- 2% in normal subjects; in all the patients, it varied from 18 to 45% and was inversely correlated (r = 0.94) to the midwall radius-to-wall thickness ratio (R/h) of the ventricle at end diastole. On the other hand, the ratio of the quantity of blood ejected per unit of LV wall volume [DELTA-V/V-omega] varied from 0.20 to 1.20 (normal subjects 0.83 +/- 0.11) and was directly correlated (r = 0.94) to R/h; using independent data in the literature, we also found a similar relationship (r = 0.80) between the ratio of quantity of blood ejected per unit of LV mass (]DELTA-V/M-omega[) and R/h. Patients with presumably abnormal myocardial function did not satisfy the relationship between [DELTA-V/V-omega[ or [DELTA-V/M-omega[ and R/h. These results show that the varying contributions of wall thickening and shortening to LV ejection as well as the efficiency in ejection of the myocardium are closely related to the geometry of the ventricle, as expressed by its R/h; for instance, patients with concentric LV hypertrophy have increase in the contribution of thickening and a decrease in the efficiency of ejection that can be related to R/h, whereas patients with abnormal myocardial function have a still lower efficiency unrelated to R/h. These results provide some insights into the mechanisms of adaptation occurring during LV hypertrophy and the relative influences of geometry and contractile function on LV ejection parameters.