Background:Non-alcoholic fatty liver disease (NAFLD) is a common liver disease in the western population and expanding disease in the world. Pathological changes in fatty liver are like alcohol liver damage, which can lead to end-stage liver disease. The prevalence of NAFLD in obese or overweight people is higher than general population, and it seems that people with high Body Mass Index (BMI) or abnormality in some laboratory tests are more susceptible for severe fatty liver and high grade of NAFLD in ultrasonography (U.S).Objectives:This study aimed to evaluate the correlation of BMI and laboratory tests with NAFLD in ultrasonography.Materials and Methods:During a multi-step process, we selected two-hundred and thirteen cases from four hundred and eighteen patients with NAFLD. Laboratory tests performed included: ALT, AST, FBS, Triglyceride and cholesterol levels, hepatitis B surface antigen, hepatitis C antibody, ceruloplasmin, serum iron, TIBC, transferrin saturation, ferritin, AMA, ANA, ANTI LKM1, serum protein electrophoresis, TSH, anti TTG (IgA). BMI and ultrasonography for 213 patients were performed, and then data was analyzed. These parameters and grades of ultrasonography were compared with the values obtained using one way ANOVA. An ordinal logistic regression model was used to estimate the probability of ultrasonography grade. The Statistical Package for the Social Science program (SPSS, version 16.0) was used for data analysis.Results:Two-hundred and thirteen cases including 140 male and 73 female, were studied. In general, 72.3% of patients were overweight and obese. Post-hoc tests showed that only BMI (P < 0.001) and TG (P < 0.011) among variables had statistically significant associations with ultrasonography grade (USG), and ordinal logistic regression model showed that BMI and AST were the best predictors.Discussion:Our results suggest that in patients with NAFLD, BMI and TG are most effective factors in severity of fatty liver disease and ultrasonography grade (USG). On the other hand, BMI as a predictor can be helpful. But, AST has not been a reliable finding, because it changes in many conditions.