We have evaluated the distribution of fatty infiltration in the liver for determination of a suitable biopsy site for diagnosis of tamoxifen-induced non-alcoholic steatohepatitis (NASH) in patients with breast cancer. Thirty-eight consecutive breast cancer patients undergoing tamoxifen treatment were analyzed by CT to identify hepatic steatosis (HS) via calculation of the liver/spleen CT ratio in Couinaud's 8 areas. We defined hepatic fatty infiltration as a liver/spleen ratio of less than 0.9. The extent and distribution of the fatty infiltration was assessed using the liver/spleen ratio of the patients who had the lowest CT ratio below 0.9 in the 8 areas. Thirteen (34.2%) of the 38 patients had hepatic fatty infiltration. The liver/spleen ratios of each area differed significantly in all patients (p<0.0001). The CT ratio of these 13 patients was significantly lower in the right lobe than the left lobe (p<0.0001), although the ratios did not differ significantly among the 4 areas of the right lobe (p=0.52). Needle biopsy for diagnosis of NASH should be performed at the right lobe, which contains significantly more infiltrated fat than the left lobe in the liver.