Objective: Dyslipidemia is a major risk factor for cardiovascular disease. Whole exome sequencing (WES) enables us to identify functionally important genetic variant and explore their association with various disease or quantitative traits. The aim of this study was to identify functional genetic variants associated with lipid traits (LDL-cholesterol, HDL-cholesterol and Triglyceride) in Koreans. Methods: We performed a two staged WES and exome genotyping array analysis. In the first stage, we performed WES in 917 subjects (619 type 2 diabetes and 298 controls). Variants that showed nominal significance of < 0.05 for association with either LDL-, HDL-cholesterol, or TG were further genotyped in the second stage analysis. In the second stage, we used semi-customized exome genotyping array to genotype 3,047 additional subjects. Genetic association was tested with linear regression analysis adjusting for age, sex, body mass index, and type 2 diabetes status using EPACTS software. Both stage 1 and 2 analyses were tested independently and meta-analysis was performed using METAL software. Results: We identified 31 SNPs (9 nonsynonymous SNPs) associated with HDL-cholesterol, 54 SNPs (22 nonsynonymous SNPs) associated with triglyceride and 18 SNPs (6 nonsynonymous SNPs) associated with LDL-cholesterol with nominally significance of P < 1.0 × 10-4. A nonsynonymous variant in APOA5 (rs2075291) had the strongest association with HDL-cholesterol (P = 2.79 × 10-12) and triglyceride (P = 1.32x10-27). A nonsynonymous variant in SORCS2 (rs537500554, P = 2.41 × 10-5) had the strongest association with LDL-cholesterol in Koreans and this SNP is the novel genetic variation in SORCS2 associated with lipid traits. Conclusion: Our study showed significant association between several nonsynonymous genetic variants with plasma lipid concentrations. Further analysis to confirm our finding is required especially for rare variants that requires even larger sample size.