Analysis of beat to beat changes in left ventricular (LV) ejection time during cardiac pacing was utilized to assess the atrial contribution to ventricular filling in 30 consecutive patients undergoing diagnostic cardiac catheterization. The group consisted of 9 normal subjects, 18 with coronary artery disease and 3 with congestive cardiomyopathy. The recordings of aortic pressure were made during atrial pacing and ventricular pacing at a rate 5 to 10 beats/min above each individual's sinus rhythm. During ventricular pacing, LV ejection time was the longest when an atrial contraction preceded a ventricularly paced beat by a physiologic interval and was approximately similar to that obtained during atrial pacing (maxET). When the atrial systole occurred with or followed the ventricularly paced contraction, LV ejection time was decreased (minET). Since maxET occurred in the presence of an effective atrial contraction to ventricular filling and minET in the absence of this contraction, the atrial contribution to ventricular filling was calculated as (maxET-minET)/maxET X 100 (%). LV volumes at end-systole (V1), before atrial contraction (V2) and at end-diastole (V3) were obtained according to the area-length method by tracing the silhouette of left ventriculograms using a computer system. The atrial contribution was calculated from LV volumes using the formula (V3-V2)/(V3-V1) X 100 (%). There was a good correlation (r = 0.88) between the atrial contributions calculated from LV ejection times and those calculated from LV volumes. In the patients with coronary artery disease and with congestive cardiomyopathy, the atrial contribution was significantly greater than in the normal subjects. The measurement of LV ejection time during ventricular pacing may be a clinically useful screening procedure to identify patients in whom physiologic pacing may be indicated.