PURPOSE: To determine the relative value of hepatic arterial and portal venous phase helical computed tomographic (CT) scans for tumor detection and vascular opacification in patients with pancreatic malignancy. MATERIALS AND METHODS: Ninety-five patients who had or were suspected of having pancreatic disease underwent dual-phase helical CT. Arterial phase scans were acquired 20-40 seconds after contrast material administration; venous phase scans, 70-100 seconds after administration. Three readers independently scored images in a blinded fashion for the presence of tumor, for lesion attenuation relative to normal pancreas, and for vascular opacification. RESULTS: The final diagnosis was pancreatic malignancy (n = 60), acute or chronic pancreatitis (n = 22), and normal pancreas (n = 13). The readers identified possible or definite tumors on arterial phase studies in 47-50 patients and on venous phase studies in 48-53 patients (P > .10). There was no statistically significant difference in tumor attenuation between scans from the two phases (P > .05). Agreement between the readers for tumor detection was not affected by the scanning phase (P > .10). Opacification of arteries and of veins was greater on arterial phase scans and on venous phase scans, respectively (P < .001). CONCLUSION: The acquisition of arterial phase scans in addition to venous phase scans does not result in improved detection of pancreatic malignancies.