The influence of apolipoprotein E polymorphism and apoE level on serum lipids and apolipoproteins was investigated in 71 healthy people and 43 patients with coronary artery disease from Shanghai. The frequency of apoE alleles was 0.06 for epsilon2, 0.86 for epsilon3, and 0.07 for epsilon4 in the healthy group, and 0.14 for epsilon2, 0.77 for epsilon3, and 0.09 for epsilon4 in the coronary artery disease group. There was no significant difference in the frequency of apoE alleles between these two groups. Serum levels of triglyceride and apo AI did not differ according to apoE genotypes, whereas serum level of apoB was significantly different according to apoE genotypes (p<0.05) both in healthy and coronary artery disease groups. However, in the healthy group, apo epsilon2 allele carriers had significantly higher level of apoE than apo epsilon3 and epsilon4 allele carriers (p<0.001) and apo epsilon4 allele carriers had significantly higher level of total cholesterol than apo epsilon3 and epsilon2 allele carriers. These were not observed in the coronary artery disease group. ApoE concentration was positively correlated with cholesterol, apoAI, and apoB levels in the control subjects and no significant correlation was observed with triglyceride level. In contrast, apoE level was positively related only to triglyceride level in the coronary artery disease group. In the control group, apoE genotypes and apoE level explained together 19.3% and 26.6% of the variability of apoB and cholesterol level, respectively, apoE polymorphism explained 23% of the variability of apoE level and apoE level explained 13.2% of the variability of apoAI level. In the coronary artery disease group, only apoE level explained 41.7% of triglyceride variability. Finally we compared our results with those previously obtained in a French healthy population, the Stanislas cohort. Results suggested that there were some difference between the Chinese control and the French subjects.