OBJECTIVE To study the etiological spectrum of acute liver failure in infants and young children and to identify clinical and biochemical markers for metabolic liver disease (MLD). METHODS This study was conducted at Department of Pediatric Hepatology, in a tertiary care specialized centre for liver diseases. All children less than 3 y of age, with liver dysfunction and INR ≥2 were included in the study. They were managed as per the departmental protocol. Included children were divided based on the etiology into 2 groups: MLD and non MLD group. Comparison analysis (MLD vs. non MLD) of the clinical and biochemical parameters was done. RESULTS There were 30 children under 3 y of age with acute liver failure (ALF) with median age of 12.5 mo. Fifteen children were less than 12 mo. MLD (33 %) and hemophagocytic lymphohistiocytosis (HLH) (17 %) together accounted for half of the cases of ALF in children below 3 y of age. The other common etiologies were drug induced liver injury and acute viral hepatitis A. Etiology remained indeterminate in 3 cases (10 %). Comparative analysis of the clinical and biochemical parameters between MLD and non MLD group showed significant difference between the two groups in the median values of age (p = 0.014), bilirubin (p = 0.017), jaundice to encephalopathy (JE) interval (p = 0.039) and blood sugar (p = 0.001). Suggestive family history (OR 3.73, 95 %CI 1.67-8.30), developmental delay (OR 4.4 95 %CI 2.03-9.51), presence of diarrhea/vomiting (OR 3.28, 95 %CI 1.32-8.13) in the history and presence of urinary non glucose reducing substance (NGRS) (OR 15.5, 95 %CI 2.26-106.87) were also significantly associated with MLD group. Only 40 % children survived with native liver. CONCLUSIONS MLD and HLH account for majority of ALF in infants. About 10 % of cases remain indeterminate. Viral hepatitis is more common in young children. Apart from clinical indicators, young age, high bilirubin, synthetic dysfunction, low sugar and NGRS in urine indicate MLD as a cause. Survival with native liver is low.