Serological surveys in many communities show a high prevalence of antibodies to hepatitis A virus (HAV) in people over the age of 50 years. However, few of that age can recall a previous episode of hepatitis, indicating that subclinical infections are common. The outcome of infection with HAV depends on the age at which infection occurs and, perhaps, the infectious dose. Fulminant disease is well recorded, with the frequency varying from one to eight per 1000 cases. Information on the frequency of hepatitis A can be obtained by analysing hospital records and notifications to health authorities or by serological surveys. In many countries, these data are limited and seriously underestimate the true.frequency of the disease. At a conservative estimate, the incidence of disease in most developed countries is probably four to five times higher than the number of notifications. HAV appears to circulate in most parts of the world and to be responsible for both epidemic and sporadic disease. Three major patterns of infection are known which reflect different epidemiological situations. These are demonstrated by different patterns of the age-specific prevalence of antibodies to HAV which reflect standards of hygiene and sanitation, the degree of crowding of the population and opportunities for the virus to survive and spread.