Between April 1984 and January 1985, in the Italian seaport of Livorno, the annual incidence of serologically confirmed acute hepatitis A doubled to 46 per 100,000 population. The exposure histories of each of 75 jaundiced subjects with serologically confirmed hepatitis A were compared with up to four, randomly chosen-, age-, sex-, and neighborhood-matched controls. Illness was strongly associated with consumption of raw mussels and clams within six weeks of onset of illness. When the two thirds of the subjects who had been exposed were classified according to the frequency with which they had recently consumed any type of raw shellfish, there was a clear dose-response relation. In February 1985, comprehensive control measures were introduced and the annual incidence of hepatitis A fell to 2.3 per 100,000 population, a 10-fold decrease from the preepidemic period.