Two populations with high risk primary liver cancer (PLC), one of 226 cases with HBsAg carriers and another of 3849 first-relatives in the pedigree with high incidence of PLC, were randomly divided into the supplementing selenium group (selenium yeast 200 g Se1Tab/day) and the control group (common yeast 1 Tab/day), and were followed-up for four years and two years respectively. In the population with HBsAg carriers, no liver cancer occurred in the supplementing selenium group; where as the liver cancer incidence rate was 1573.03/10(5) in the control group. Among the first relatives. The liver cancer incidence rate in the supplementing selenium group was 219.37/10(5); and 553.15/10(5) in the control group. The results showed that the incidence of PLC in the supplementing selenium group was significantly lower than in the control group. This study indicates that selenium has distinct anti-PLC effect.