PURPOSE: To evaluate dynamic CT features and its clinical courses of eosinophilic hepatic abscess. MATERIALS AND METHODS: Two-phase dynamic CT findings and the clinical courses of 13 pathologically proven cases of eosinophilic abscess were reviewed. All patients showed peripheral eosinophilia, and diagnoses were confirmed by ultrasound-guided biopsy(n=9) or operation(n=4). In two of the four patients who underwent segmental hepatectomy, worms of the species Fasciola hepatica were detected. Follow-up CT scans after treatment with antibiotics or praziquantel were available in seven and eight patients, respectively. RESULT: All hepatic lesions were found ina subcapsular location or in contact with Glisson's capsule around the bile duct. Arterial-dominant phase CT(n=11)demonstrated clusters of ill-defined low density masses without rim enhancement. Late-phase CT(n=13) more clearly depicted clustering lesions with enhancing rims and diminution of the low-density area. Follow-up CT scans aftertreatment with antibiotics(n=7) showed no change in the lesions in three patients and slight shrinkage of the mainmass with additional new lesions in four. On CT scans of nine patients performad after praziquantel therapy, hepatic masses were seen in all patients to be very slightly smaller after improvement of peripheral hypereosinophilia. CONCLUSION: Two-phase dynamic CT features appear to be helpful for the diagnosis ofeosinophilic hepatic abscess in patients with peripheral eosinophilia. Parasitic infestation by Fasciola hepaticafor example, is the presumed cause of such abscesses, though further studies are required.