To determine the effect of contrast medium injection rate and biphasic versus uniphasic bolus on hepatic parenchymal enhancement at computed tomography (CT), 105 patients were randomized into four protocols. Hepatic and aortic enhancement curves were measured for each protocol. The optimal scanning interval (the duration between onset of a desired hepatic enhancement threshold and decline of enhancement below the threshold or onset of the equilibrium phase) and contrast enhancement index (CEI) (area under the hepatic enhancement curve above a desired threshold during the nonequilibrium phase) were the primary determinants of merit. Measurements were performed at hepatic enhancement thresholds between 10 and 60 HU. At most thresholds of hepatic enhancement, optimal scanning intervals were significantly longer and CEIs were significantly higher for the biphasic protocols than for the uniphasic protocols (P < .03). A biphasic injection with a high initial flow rate produces high peak contrast enhancement but delayed onset of equilibrium and thus provides a longer optimal scanning interval.