Peroxisome proliferator activated receptors (PPARs) are transcription factors involved in the regulation of key metabolic pathways. Numerous in vivo and in vitro studies have established their important roles in lipid metabolism. A few SNPs in PPAR genes have been reported to be associated with lipid levels. In this study, we aimed to investigate the interactive effects between single nucleotide polymorphisms (SNPs) in three PPAR isoforms α/δ/γ and other genetic variants across the genome on plasma high-density lipoprotein-cholesterol (HDL-C) levels. Study subjects (N = 2003) were genotyped using Illumina HumanOmniZhongHua-8 Beadchip. Fifty-three tag SNPs ± 100 kb of PPAR α, δ, and γ (r2 < 0.2) were selected. The effect of interactions between PPAR SNPs and those across the genome on HDL-C was tested using linear regression models. One statistically significant interaction influencing HDL-C was detected between PPARδ SNP rs2267668 and epithelial membrane protein 2 (EMP2) downstream SNP rs7191411 (N = 1993, β = 0.74, adjusted P = 0.022). This interaction was successfully replicated in the meta-analysis of two additional Chinese cohorts (N = 3948, P = 0.01). The present study showed a novel SNP × SNP interaction between rs2267668 in PPARδ and rs7191411 in EMP2 that has significant impact on circulating HDL-C levels in the Singaporean Chinese population.