Purpose: Several reports have suggested the development of hepatic steatosis after gastrointestinal bypass surgery in patients without metabolic syndrome. The aim of his study was tried to investigate the effect of pancreaticoduodenectomy (PD) on the course of hepatic steatosis by use of CT attenuation values.Materials and Methods: A retrospective case-crossover study by using the PD database of NCKUH. Between January 2004 and January 2008, patients admitted for PD were enrolled. Exclusion criteria included multiple liver metastasis, severe obesity (BMI>30), unavailable pre-op and post-op over 6-months CT images. The pre-PD and post-PD liver attenuation, ratio of liver-to-spleen attenuation, and pre-PD and post-PD difference between liver and spleen attenuation were compared. Paired student t test was used for statistical analysis.Results: 99 patients were extracted from the database and finally 50 patients were eligible. The mean follow-up period for the study group was 18.0±1.6 month, range from 6 to 43 months. The pre-PD and post-PD liver attenuation was 52.3± 1.1H and 47.6± 2.0 H (p=0.044). The pre-PD and post-PD ratio of liver-to-spleen attenuation were 1.12±0.02 and 1.01±0.04 (p=0.033). The pre-PD and post-PD difference between liver and spleen attenuation were 5.52± 1.07H and 0.55±1.85 H (p=0.031). Conclusion: According to the quantitative analyses of attenuation on CT image, the progression of hepatic steatosis was found after pancreaticoduodenectomy. The mechanism of post-PD hepatic steatosis is necessary to be further elucidated.