During an outbreak of hepatitis A that occurred in Gloucester, UK between September 1989 and January 1991, 162 clinical cases were identified through notifications and laboratory reports, a monthly attack rate of 1.05 per 10,000 residents. The highest attack rate was seen in 5-14-year-olds. There were significant correlations between hepatitis A attack rates in the electoral wards of Gloucester and with the Jarman UPA 8 scores for the wards and with overcrowding, unemployment, under 5-year-olds and ethnic minority. The use of human normal immune globulin prophylaxis (HNIG) for household contacts was unsuccessful in ending the outbreak, partly because only one third of cases reported a household contact with recent hepatitis A. Our experience does not support the use of HNIG in stopping community-wide outbreaks of hepatitis A. Two public health campaigns were mounted during the outbreak; both were followed by a fall in the number of cases. Greater priority should be given to the implementation and evaluation of public health campaigns in future community-wide outbreaks of hepatitis A.