Purpose: Obesity is implicated in the severity and outcome of acute pancreatitis. The role of regional distribution of fat in the severity and outcome of acute pancreatitis (AP) has not been studied. We aim to study the relationship of abdominal adiposity and hepatic steatosis with the severity and outcome of acute pancreatitis. Methods: A retrospective chart review of 108 (mean age 53 yrs, male 52%) of 257 patients with AP seen between January 2002 and December 2009 meeting the inclusion criteria was performed. Ranson's score and CT severity index calculations were performed. Hepatic attenuation index (HAI) was obtained by calculating the ratio of hepatic to splenic attenuation (Hounsfield units) on an unenhanced CT; HAI <0.8 was used to diagnosis moderate to severe macrovesicular steatosis. Total abdominal fat was calculated using sagital and coronal reformation images on CT of the abdomen and pelvis using data from the superior end plate of T10 to the inferior symphysis pubis. Hounsfield units ≤ 200 HU and ≤20 HU were excluded yielding total abdominal fat content; a histogram profile of this remaining volume using a post-processing Workstation (GE AW 4.2, GE Milwaukee, WI) was used to calculate the fat volume. Results: Etiology of AP was alcohol 19(17.6%), ERCP-induced 8(7.6%), gallstone 42(39%) or other etiologies 39(36%). HAI < 0.8 was noted in 23(21.3%) patients. The severity of pancreatitis by Ranson's criteria at 48 hrs (1.91±1.65vs. 1.20±1.26, p=0.04) and by CT severity index criteria (4±3.2vs. 2.2±1.6, p=0.04) were significantly higher in patients with low-HAI (fatty liver) compared to high-HAI (non- fatty liver). Patients with low HAI had longer hospital stay (13.8±11.2 vs. 11.5±17, P=0.02), more frequent ICU admission (47.8 vs. 24.7, P=0.03), more frequent need for antibiotics (69.6 vs. 30.6, P=0.001) and higher in-hospital mortality (13% vs. 1.2%, P= 0.008) compared to patients with higher HAI. Although higher number of patients also had CT evidence of necrosis (21.7% vs. 10.6%, P=0.17) and pseudocyst formation (13% vs. 7.1%, P=0.39), this did not reached statistical significance. Mean total abdominal fat content was not significantly different in patients with severe (score > 3) vs. milder (score < 3) pancreatitis by Ranson's criteria (15.6±6.7 vs. 13±4.7 liters, p=0.14). Conclusion: Lower hepatic attenuation index, a measure for hepatic steatosis on unenhanced CT scan, is a novel non-invasive marker for severe outcomes in patients with acute pancreatitis.