The percentage of shortening of the echocardiographic left ventricular dimension (% delta D) was prospectively evaluated in 42 patients without detectable asynergy during diagnostic cardiac catheterization and was found to correlate well with angiographic ejection fraction (r = 0.90). Ejection fraction was calculated as the product of % delta D X 1.7 or as % delta (D2), both formulae having similar degrees of accuracy and a better correlation with the angiographic determination than conventional formulae. Ejection fractions (angiographic and echocardiographic) of 51 percent or greater were always associated with a % delta D of 30 percent or more. In five patients the echocardiographically derived ejection fractions were normal (greater than or equal to 51 percent), while the angiographic ejection fractions were reduced; four of these patients had valvular regurgitation. End-diastolic volumes were calculated from end-diastolic echocardiographic dimensions utilizing a linear regression equation derived from correlating the end-diastolic echocardiographic dimension with the end-diastolic volume in 27 patients without valvular regurgitation (end-diastolic echocardiographic dimension ranged from 3.7 to 8.2 cm). The value for stroke volume determined as the product of calculated end-diastolic volume times ejection fraction correlated with the angiographically determined stroke volume (r = 0.88; standard error of estimate, +/- 11 ml) better than the value for stroke volume derived from conventional echocardiographic formulae.