OBJECTIVE We hypothesized that CT with arterial portography (CTAP) might depict tumors undetected by CT with infusion hepatic arteriography (CTIHA), and the converse. Accordingly, we compared the detection rate of malignant tumors using CTAP, CTIHA, and simultaneous use of both techniques to determine whether CTIHA can be used to distinguish malignant from benign hepatic nodules. METHODS Fifty-four candidates for partial liver resection who underwent combined CTAP and CTIHA were studied. Histopathologic study confirmed 77 malignant and 15 benign nodules in these patients. CTAP, CTIHA, and combined CTAP and CTIHA were used by three observers before surgery, and the sensitivities of the three methods in detecting malignant tumors were compared. Contrast-enhancement patterns of the nodules on CTIHA scans were determined. RESULTS Sixty-eight malignant nodules (88%) were detected with CTAP, 64 (83%) with CTIHA, and 69 (90%) with simultaneous interpretation of both scans. These differences were not significant. CTIHA showed 16 malignant and 12 benign small nodules. Of these, 13 malignant and two benign nodules had rim enhancement. Use of rim enhancement as the criterion for malignancy of these small nodules gave an accuracy rate of 82%. CONCLUSIONS CTAP, CTIHA, and simultaneous use of both techniques are similar in their ability to detect malignant hepatic tumors. CTAP alone is recommended for the detection of malignant hepatic tumors. CTIHA is of value in differentiating malignant from benign small nodules.