Systolic time intervals derived from Doppler velocimetry measurements were used instead of direct pulmonary to systemic flow ratio measurements in adults with atrial septal defect to quantify left to right atrial shunts. Thirteen normal subjects and 25 patients with uncomplicated atrial septal defect confirmed by cardiac catheterisation were studied. The pulmonary to systemic flow ratio (Qp:Qs) expressing the shunt size was determined by the Fick method; in normal subjects the Qp:Qs ratio was assumed to be equal to 1.0. The pulsed Doppler analogue velocity recording of flow in the pulmonary artery and the ascending aorta was taken as indicating the ejection time of each ventricle and the Q wave of the electrocardiogram as indicating the onset of systole. From these measurements the ratios of the pre-ejection periods to the ejection times (haemodynamic ratio) were calculated for each ventricle and the ratios of each variable (pre-ejection period, ejection time, and haemodynamic ratio) were calculated for both ventricles. Significant differences were found between the normal subjects and the patients with atrial septal defect for all these ratios. When the Doppler findings and the Fick measurements of Qp:Qs were compared the best linear correlation coefficient was for the left to right haemodynamic ratio. It is concluded that the use of a ratio involving several variables, such as the pre-ejection period and the ejection time for both ventricles, improves the reliability of this method, which appears to be applicable in adults.