In the rapidly developing city of Almaty, Kazakhstan, rates of hepatitis A have fallen, but no data on prevalence of antibody to hepatitis A virus (anti-HAV) exist with which to interpret incidence data. In the autumn of 2001, we determined the anti-HAV prevalence among household and school contacts of hepatitis A cases. For contacts aged 0–4 years, 5–9 years, 10–14 years, 15–19 years, or 20–30 years, immune prevalences were 9, 12, 33, 33 and 77% respectively, among immediate-family household contacts and 15, 28, 49, 52 and 77% respectively, among community contacts. Child community contacts were more likely to be immune than their immediate-family household counterparts (odds ratio 2·0, 95% confidence interval 1·3–3·2). Almaty is experiencing an epidemiological shift in hepatitis A incidence. Feasible and effective prevention strategies using hepatitis A vaccine should be explored.