In 76 adult males quantitative angiographic measurements of left ventricular size were compared to 12 plain chest film measurements of heart size and shape. Of the 18 normal cases, only Rigler's B measurement correlated with left ventricular volume (r = .64). However, this correlation was not statistically significant. Of the 58 patients with isolated aortic valve disease, 19 had increased left ventricular mass (hypertrophy) and 39 had both increased left ventricular end-diastolic volume and mass (hypertrophy and dilatation). In the hypertrophy group, plain film heart volume was the most sensitive detector of an enlarged left ventricle (21% false negative rate) and correlated with left ventricular end-diastolic volume (r = .62). Using stepwise discriminant analysis, the combination of heart volume, leftness of the heart, apex position, and roundness of the left ventricle decreased the false negative rate to 5%. In the hypertrophy and dilatation group, plain film heart volume had the highest correlation with left ventricular volume and mass (r = .66) and a false negative rate of 8%. No single variable or combination of variables could usefully discriminate between the hypertrophy and hypertrophy and dilatation groups. These data support the conclusion that plain film heart volume is the best single measurement for detecting left ventricular enlargement.