Using pulsed Doppler echocardiography, left ventricular inflow flow volume (LVIV) and outflow flow volume (LVOV) were noninvasively determined, and the ratio of pulmonary to systemic flow (Qp/Qs) was evaluated as the ratio of LVIV to LVOV (LVIV/LVOV). Thirty patients with ventricular septal defect (VSD) were studied, and 47 cardiac patients without aortic or mitral valve disease or intracardiac shunt served as controls. LVOV was derived from the left ventricular ejection flow velocity and the outflow tract diameter immediately proximal to the aortic valve ring. LVIV was derived from the transmitral flow velocity and the M-mode tracing of mitral valve motion. Doppler-determined cardiac outputs (COin and COout) were calculated as the products of LVIV or LVOV as and heart rates. Cardiac outputs were also determined by the dye dilution method (COdye) references for comparison with Doppler-determined cardiac outputs. There were good correlations between COdye and COin (y = 1.18x-243, r = 0.85, p less than 0.005, SEE = 1026 ml/min) and between COdye and COout (y = 1.16x-323, r = 0.90, p less than 0.005, SEE = 639 ml/min). LVIV and LVOV correlated well in the controls (y = 0.95x + 5.3, r = 0.94, p less than 0.005, SEE = 6.6 ml). LVIV/LVOV was 0.97 +/- 0.1 (mean +/- SD) in the controls; whereas LVIV/LVOV (1.86 +/- 0.90) was significantly higher in patients with VSD (p less than 0.01) and this ratio correlated well with Qp/Qs by an oximetry (r = 0.98, SEE = 0.20, n = 14), including patients associated with pulmonary regurgitation. These findings indicate that our method permits determination of LVIV with a high degree of accuracy and that the Doppler-determined LVIV/LVOV is clinically useful to evaluate accurately the magnitude of shunt flows in patients with VSD.