To optimize findings at dynamic gadolinium-enhanced hepatic magnetic resonance imaging in the hepatic arterial phase, a timing examination was performed after injection of a 0.5-mL timing bolus of gadopentetate dime-glumine. In the experimental group (n = 28), power injection was used and the imaging delay was determined in each patient on the basis of the results of the timing examination. In the control group (n = 28), hand injection and a fixed 20-second imaging delay were used. Arterial-phase examinations (defined as relative liver enhancement of not more than 30% of peak parenchymal enhancement) were successful In 26 (93%) experimental group patients and in 17 (61%) control group patients.