Technique and errors of quantitative single plane ventriculography (SPV), and the methods of Dodge and Sandler (Dodge et al, 1962; Sandler and Dodge, 1968) and Greene et al (1967) for determining left ventricular (LV) volume by SPV were evaluated in the intact dog. Stroke volume (SV) and cardiac output (Q) by ventriculography were compared with those obtained by the Fick and dye dilution methods, and their combination. The end-diastolic (EDV) and end-systolic volume (ESV) corrected by appropriate regression equations, the SV (SV=EDV - ESV), and the ejection fraction (EF) were: Dodge and Sandler, 2.26 1.35, 0.91 ml/kg, and 0.42; Greene et al, 2.41, 1.50, 0.91 ml/kg, and 0.39. Using adequate methods, accurate correction for x-ray image magnification, and the appropriate regression equations the standard error of LV volume calculation by single plane radiography, due to unavoidable technical inaccuracies, should not exceed 10% for a single measurement and 5% for the average of two or more successive cardiac cycles. The systematic overestimation of LV volume inherent in ventriculography, and caused by the addition of contrast medium, should not exceed 5% of the true value.