In January 2004, we investigated a cluster of acute hepatitis in Baripada, Orissa, India. Between January and March 2004, 538 cases (definition: fever with loss of appetite and jaundice) were reported (attack rate: 263 per 100 000, 5 deaths, case fatality rate: 0.93%). Forty-seven of 48 sera were positive for IgM antibodies to hepatitis E virus. Cases peaked in February 2004 and decreased rapidly, suggesting a common source outbreak. Five neighbourhoods supplied by a common water supply were most affected. Ninety-one percent of the 538 cases and 30% of 538 unaffected controls reported drinking water from one source (odds ratio 31, 95% CI 27-48). The neighbourhood's water was pumped directly from a river and had not been treated during a 10-day period in early January (1 month before the peak of the outbreak) because of a strike at the treatment plant. This large hepatitis E outbreak was associated with drinking untreated raw river water. Measures must be in place to check the quality of municipal water and to ensure essential services in case of strikes.