The hepatotoxic effect of arsenic when used in therapeutic dose has long been recognized. We described the nature and degree of liver involvement and its pathogenesis due to prolonged drinking of arsenic-contaminated water in West Bengal, India. From hospital-based studies on 248 cases of arsenicosis, hepatomegaly was found in 190 patients (76.6%). Non cirrhotic portal fibrosis was the predominant lesions in 63 out of 69 cases who underwent liver biopsy. The portal fibrosis was characterized by expansion of portal zones with streaky fibrosis, a few of which contained leash of vessels. However, portal hypertension was found in smaller number of cases. A cross-sectional epidemiological study was carried out on 7683 people residing in arsenic-affected districts of West Bengal. Out of these, 3467 and 4216 people consumed water-containing arsenic below and above 0.05 mg/l, respectively. Prevalence of hepatomegaly was significantly higher in arsenic-exposed people (10.2%) compared to controls (2.99%, P < 0.001). The incidence of hepatomegaly was found to have a linear relationship proportionate to increasing exposure of arsenic in drinking water in both sexes (P < 0.001). In an experimental study, BALB/C mice were given water contaminated with arsenic (3.2 mg/l) ad libitum for 15 months, the animals being sacrificed at 3-month intervals. We observed progressive reduction of hepatic glutathione and enzymes of anti-oxidative defense system associated with lipid peroxidation. Liver histology showed fatty infiltration at 12 months and hepatic fibrosis at 15 months. Our studies show that prolong drinking of arsenic-contaminated water is associated with hepatomegaly. Predominant lesion of hepatic fibrosis appears to be caused by arsenic induced oxystress.