Some of them showed robust CAD association, SNP association showed inconsistent magnitudes of CAD association. It is assumed that such inconsistency is explained by pleiotropic actions of the lipid loci in part. In the present study, to test associations between lipid traits and CAD for SNPs from 22 candidate loci recently reported by GWA studies and their meta-analyses in populations of European descent, we performed a replication study in Japanese populations. From the viewpoint of population genetics, generalization of lipid association results previously identified in European GWA studies to non-European populations is an issue of interest, because it can facilitate the fine mapping of common causal variants by providing clues to whether SNPs identified in European GWA studies are simply tag-SNP or “synthetic association” markers or are more likely to be true functional variants. Thus far, only a few studies have addressed this in non-European populations, and our study is the first replication study in east Asians, with focus on both lipid trait and CAD associations. Replicating a study of candidate loci previously identified by GWA meta-analyses of European-descent populations, we have found some degree of ethnic diversity in lipid variants, while 18 of 22 tested loci are associated with lipid traits in the Japanese. The loci showing strong lipid associations were in good agreement between the ethnic groups except APOB, where associations with LDL-C and TG were relatively weak in the Japanese. Also, in the present study, we confirmed significant genetic impacts of 4 loci–SORT1, APOA5, LDLR, and APOE–on CAD in the Japanese. Of note, the effect size of APOE variants on CAD was significantly large in the Japanese. Moreover, it is of interest that as compared with the results for Europeans, the variance for LDL-C levels explained by individual SNP loci tended to be smaller in the Japanese, despite an overall cross-population consistency of genetic variants. For the most significant locus, APOE, 3 major isoforms are known to exist; E2 and E4 isoforms can be differentiated from E3 by rs7412 and rs429358, respectively. In agreement with a previous study of meta-analysis, E2 and E4 exerted GANT61 decreasing and increasing genetic effects on LDL-C, respectively, as compared to E3. While E2 carriers had a significantly decreased risk of CAD, E4 carriers showed no increased risk of CAD in the Japanese population, which is inconsistent with the previous reports. In addition, we found that effect sizes of APOE variants were heterogeneous between the current study and those previously reported. The genetic impacts on CAD seem to be more prominent for APOE rs7412 than the loci that we previously detected in the Japanese GWA scan. Because rs7412 and rs429358 themselves and their proxies are not included in the list of SNPs that are assayed by most of the GWA scan platforms, it is likely that these SNPs have failed to be tested for CAD associations in the previous GWA studies. As an approach to examining clinical relevance of lipidassociated SNPs, we assessed whether they are also associated with CAD in a manner consistent with established epidemiological relationships; i.e., SNP alleles that increase LDL-C or TG or that decrease HDL-C should be associated with increased risk of CAD, in proportion to the genetic effects on lipid traits. We inspected correlations between genetic effects on CAD risk and those on lipid traits.