To prospectively evaluate the safety and efficacy of combined unenhanced and gadoxetic acid disodium (Gd-EOB-DTPA)-enhanced magnetic resonance (MR) imaging compared with unenhanced MR imaging and triphasic contrast-enhanced spiral computed tomography (CT) for the detection and characterization of focal liver lesions. The study was reviewed and approved by the institutional review board at each of the 15 centers involved in the study, and informed written consent was given by all patients. In total, 178 patients with suspected focal hepatic lesions (based, in most patients, on CT, tumor marker and ultrasound examinations) underwent combined MR imaging with a single, rapid injection of Gd-EOB-DTPA 0.025 mmol/kg, including T1-weighted dynamic and delayed MR images 20 to 40 minutes postinjection. Triphasic contrast-enhanced CT, the comparator examination, was performed within 4 weeks of MR imaging. Standard of references (SOR) were resection histopathology and intraoperative ultrasonography, or combined CT during arterial portography and CT hepatic arteriography; in cases where, although the major lesions were treated, some lesion(s) were not treated, follow-up superparamagnetic iron oxide-enhanced MR imaging was additionally performed. All images were assessed for differences in lesion detection and characterization (specific lesion type) by on-site readers and 3, blinded (off-site) reviewers. All adverse events (AEs) occurring within 72 hours after Gd-EOB-DTPA administration were reported. Overall, 9.6% of patients who received Gd-EOB-DTPA reported 21 drug-related AEs. A total of 151 patients were included in the efficacy analysis. Combined MR imaging showed statistically higher sensitivity in lesion detection (67.5%-79.5%) than unenhanced MR imaging (46.5%-59.1%; P < 0.05 for all). Combined MR imaging also showed higher sensitivity in lesion detection than CT (61.1%-73.0%), with the results being statistically significant (P < 0.05) for on-site readers and 2 of 3 blinded readers. Higher sensitivity in lesion detection with combined MR imaging compared with CT was also clearly demonstrated in the following subgroups: lesions with a diameter <or=20 mm (lesions <or=10 mm: 38.0%-55.4% vs. 26.1%-47.3%, respectively; lesions 10-20 mm: 71.1%-87.3% vs. 65.7%-78.4%, respectively); in cirrhotic patients (64.5%-75.4% vs. 54.5%-70.3%, respectively); and in patients with hepatocellular carcinoma (66.6%-78.6% vs. 59.1%-71.6%, respectively). Combined MR imaging demonstrated a higher proportion of correctly characterized lesions (50.5%-72.1%) than unenhanced MR imaging (30.2%-50.0%; P < 0.05 for all), whereas there were no significant differences compared with CT (49.0%-68.1%), except for one blinded reader (P < 0.05). In this study, hepatocyte-specific Gd-EOB-DTPA was shown to be safe and to improve the detection and characterization of focal hepatic lesions compared with unenhanced MR imaging. When compared with spiral CT, Gd-EOB-DTPA enhanced MRI seems to be beneficial especially for the detection for smaller lesions or hepatocellular carcinoma underlying cirrhotic liver.