Ventriculography was performed with conventional biplane fluoroscopic equipment; a 1:4 dilution of contrast medium in a standard injectate volume was delivered by a power injector. The biplane images were acquired with two video tape recorders and enhanced by means of a portable digital image processor. By a simple modification and calibration of our standard light pen-computer system, volumes could be calculated and global and segmental wall motion analysis performed. Fifteen patients underwent conventional and low-dose left ventriculography. Good correlation was found in end-diastolic volume (EDV), in end-systolic volume (ESV), and in ejection fraction (EF) (r = 0.90, r = 0.93, and r = 0.92, respectively). The quality of the images enabled a frame-by-frame analysis of a cardiac cycle to be performed (r = 0.99 when compared in one patient) and regional wall motion abnormalities identified (21 out of 23 abnormal segments). Although conventional ventriculography caused a rise in left ventricular end-diastolic pressure (17 +/- 8 vs 23 +/- 11, p less than 0.002) and no change in systolic blood pressure (132 +/- 23 vs 135 +/- 23, p greater than 0.5), low-dose ventriculography had no effect on left ventricular end-diastolic pressure (18 +/- 7 vs 21 +/- 10, p greater than 0.05) and produced a rise in systolic blood pressure (144 +/- 20 vs 149 +/- 22, p less than 0.05). A system of digital enhancement that is capable of producing high-quality ventriculograms with lessened contrast morbidity can be assembled with minimal expense and adapted to any conventional catheterization laboratory.