It is still unclear whether Super Paramagnetic Iron Oxide-Magnetic Resonance Imaging (SPIO-MRI) is a clinically useful imaging modality for patients with hepatocellular carcinoma (HCC). This study searched for the clinical usefulness and limitations of SPIO-MRI with respect to tumor detection capacity, false positive and negative rate, and early recurrence rate.From December 1999 to February 2001, 218 patients who were surgical candidates by 3-phase dynamic helical CT (3dHCT) were enrolled. We reviewed the medical records and radiologic findings, retrospectively, and postulated the post-operative pathologic findings and the early recurrences within 3 months as the standards for the true positive lesion.The mean number of nodules detected by SPIO-MRI was significantly more numerous than that of 3dHCT (p<0.01). Modifications of treatment strategy due to the discordant findings between SPIO-MRI and 3dHCT for tumor resectability were observed in 22 (10.1%) out of 218 patients. Early recurrences were observed in 10 patients (7.8%). The false positive and negative rates of SPIO-MRI were 6.3% and 13.3%, respectively.We demonstrated that the tumor detection rate of SPIO-MRI was better than that of 3dHCT. Given the relatively acceptable false positive and negative rates, SPIO-MRI could be an appropriate preoperative imaging modality for patients with HCCs.