We aimed to evaluate the long-term efficacy of the hepatitis B vaccine in China. In an initial efficacy study, participants were collected from a cluster-randomized clinical trial conducted in 1983-90 in Qidong. All the participants in the vaccination group were vaccinated at birth, 1 and 6 months of age, and no intervention was implemented to the control group. In this 37-year extended follow-up study, the Poisson regression method was employed to derive rates per 10 person-years. The frailty Cox proportional hazard regression models obtained the hazard ratio (HR). Cumulative incidence/mortality rates were calculated and compared with log-rank tests. 41,136 in the vaccination and 41,730 in the control group were recorded. The incidence rate of liver cancer was significantly lower in the vaccinated group than in the control group [HR, 0.28; 95% confidence interval (CI) 0.11-0.70, P = 0.007]. The vaccine offers 72% (95% CI, 30-89) protection to prevent the occurrence of liver cancer. There is 70% (95% CI, 23-88) protective efficacy against liver cancer deaths and 64% (95% CI, 27-82) benefits in the prevention of deaths associated with liver diseases. Hepatitis B vaccine given at birth shows excellent protective effects in preventing the development of liver cancer and reducing mortality from liver cancer and liver diseases.