Hepatic Angiography by selective catheterization of the celiac or hepatic arteries is a useful procedure to evaluate the liver for malignant disease. Bierman et al. in 1951 (1) first described in vivo hepatic angiography applied to the study of hepatic neoplasia. Unfortunately, their technic, utilizing a catheter inserted by arterial cut-down, was fraught with hazard (2) and therefore was impractical for general use except as an investigative effort in terminally ill patients. With percutaneous catheter introduction, as devised by Seldinger (3), hepatic angiography was facilitated. In 1958, Odman (4) reported 61 cases of selective celiac angiography by the percutaneous approach. This paper will present the findings of hepatic angiography in 40 cases of malignant disease drawn from a total experience of 200 cases of selective celiac, hepatic, and superior mesenteric angiography performed at the Massachusetts General Hospital, Boston, Mass. Methods and Materials Selective celiac and selective hepatic angiography was performed by the Seldinger technic of percutaneous catheter introduction, usually through the femoral route, although occasionally left brachial or left axillary approaches were utilized. A preshaped Kifa catheter was employed and 30 to 50 cc of 76 per cent methylglucarnine diatrizoate was injected under pressure. With a serial film-changer films were obtained at the rate of 2 per second for eight seconds and 1 every third second for an additional twenty seconds, in the anteroposterior projection. Frequently, oblique projections were obtained and additional injections were performed. A replaced or accessory hepatic artery originating from the superior mesenteric artery occurs in 14 per cent or more of cases (5, 6), and in such instances superior mesenteric artery catheterization is required. Results The accuracy of this technic in 71 cases is recorded in Table I. Proof was obtained by autopsy, laparotomy, or positive needle biopsy. Negative needle biopsies were not included in this number since neoplastic involvement could easily be missed by a percutaneous biopsy. Because a negative surgical exploration does not absolutely rule out the presence of tumor, no attempt to draw statistical conclusions about negative cases was made, but the results show that a negative angiogram does not necessarily exclude the presence of tumor in the liver. The accuracy of positive interpretations, however, was 92 per cent. Of 40 cases of hepatic neoplasms, 33 were proved; the remainder showed characteristic angiographic findings, and a clinical course and laboratory data were consistent with malignant. disease in the liver. Although the origin of the hepatic arteries is variable, the intrahepatic arteries normally reveal a random arborization of the left and right branches, gradually tapering in a smooth fashion and showing a uniform intrahepatic distribution of the fine branches.