Although intravenous digital subtraction ventriculography (IDSV) is increasingly used to estimate end-diastolic left ventricular volume (EDV), end-systolic left ventricular volume (ESV) and left ventricular ejection fraction (EF), its ability to reproduce the precise estimates provided by left ventricle cineangiography (LVCA) and its role in clinical cardiology have not been unequivocally established. In 32 patients subjected to cardiac catheterization for a variety of cardiac disorders and a normal or reduced left ventricular function the EDV, ESV and EF provided by a 30 degrees right anterior oblique LVCA were compared with those provided by a 30 degrees right anterior oblique IDSV. The mean EDV, ESV and EF obtained by IDSV in the 32 patients were superimposable on those obtained by LVCA. The individual EDV, ESV and EF values provided by the two methods were all related in a close linear fashion. For EF the correlation coefficient was 0.98 and the 90% confidence interval of the mean difference between the two series of values was +/- 6.1%, i.e. +/- 10% error compared to the mean EF provided by LVCA. Thus IDSV is a reliable and not too invasive method for estimating left ventricle volumes and ejection fraction. It might provide serial estimations with a better assessment of the evolution of a patient's disease and the effect of treatment.