Computed tomographic arterial portography (CTAP) has been shown to be the most sensitive preoperative test for determining resectability of hepatic lesions but we have shown it to have low specificity. Intraoperative ultrasound (IOUS) evaluation of the liver has also been proposed as an accurate means of assessing resectability. We sought to compare the effectiveness of the two modalities. Fifty-six patients who had been deemed candidates for liver resection based on CTAP findings underwent systematic exploration, liver mobilization, and IOUS examination. Ultrasound findings were compared with results of CTAP. In 46 patients the IOUS findings were in complete agreement with those of CTAP. In 10 patients CTAP lesions could not be verified by IOUS and these patients did not undergo resection. Follow-up of these 10 patients revealed eight who did not have progression of malignancy at the CTAP-predicted site (CTAP false positive). Two patients did have progression at a CTAP-positive IOUS-negative site (IOUS false negative). Sensitivity for CTAP and IOUS was 100% and 96%, respectively. Specificity for IOUS was 100%. These findings demonstrate the high sensitivity of CTAP and the high sensitivity and specificity of IOUS. CTAP may "overcall" hepatic lesions but IOUS can correctly identify these false positives in most instances. Because CTAP is useful for determining which patients might benefit from surgical exploration, we conclude that the two modalities are complementary for the assessment of resectability of hepatic lesions. The false positive rate for CTAP implies that caution must be used when declining to operate on patients on the basis of this test.