Longan County is considered a highly endemic area for hepatitis B virus (HBV). The plasma-derived vaccine has been used in newborns in this area since 1987. A cross-sectional survey was conducted to evaluate the long-term effectiveness of this vaccine. In total, 1634 participants born during 1987–1993 and who had received a series of plasma-derived HB vaccinations at ages 0, 1, and 6 months were enrolled. Serological HBV markers were detected and compared with previous survey data. Overall the prevalence of hepatitis B surface antigen (HBsAg) in all participants was 3·79%; 3·47% of subjects who had received the first dose within 24 h were HBsAg positive, and 8·41% of subjects who had received a delayed first dose were also HBsAg positive. There were 1527 subjects identified who had received the first dose within 24 h and whose HBsAg and anti-HBc prevalence increased yearly after immunization, while the anti-HBs-positive rate and vaccine effectiveness declined. The geometric mean concentration of antibody in the anti-HB-positive participants was 55·13 mIU/ml and this declined after immunization. Fewer than 2·0% of participants had anti-HB levels ⩾1000 mIU/ml. The data show that the protective efficacy of the plasma-derived vaccinations declined and administration of HB vaccine within 24 h of birth was very important. To reduce the risk of HBV infection in this highly endemic area, a booster dose might be necessary if anti-HBs levels fall below 10 mIU/ml after age 18 years. Furthermore, studies on the immune memory induced by plasma-derived HB vaccine are needed.