BACKGROUND: To date, the noninvasive diagnostic tests for hepatic fibrosis in subjects with nonalcoholic fatty liver disease (NAFLD) have proven to be suboptimal. We evaluated the validity of a recently proposed "NAFLD fibrosis score" to identify liver fibrosis in morbidly obese individuals with elevated and normal alanine aminotransferase (ALT) levels. METHODS: Medical records of 401 patients that underwent a gastric bypass operation and intraoperative liver biopsy were analyzed. Three hundred thirty one patients with biopsy-proven NAFLD were included in the study (group A). These patients were divided into two ALT groups based on their levels according to the new proposed normal range: group B elevated level (ALT > 19 U/L in females and >30 U/L in males, n = 221) and group C normal ALT (n = 110). Diagnostic accuracy of the system was assessed for the presence/absence of any fibrosis, significant fibrosis (stage 2-4), and advanced fibrosis (stages 3 and 4) in all of the groups. RESULTS: The prevalence of advanced fibrosis in our cohort was about 14%. The low NAFLD fibrosis score demonstrated high accuracy for ruling out advanced fibrosis, with negative predictive value (NPV) of 98 and 99% in groups A and B, respectively. The NPV for significant fibrosis in groups A, B, and C was 87, 88, and 88%, respectively. The respective positive predictive value for the high NAFLD fibrosis score for the presence of any fibrosis was 88, 95, and 77% in groups A, B, and C. CONCLUSIONS: The NAFLD fibrosis score may be a useful noninvasive approach for excluding significant and advanced fibrosis and in morbidly obese patients.