The objective of this paper was to identify the single nucleotide polymorphisms (SNPs) that show unshared effects on plasma triglyceride (TG) levels and to investigate whether these SNPs show statistically independent effects on plasma TG levels. In total, 59 polymorphisms in 20 genes involved in lipid metabolism were investigated. Polymorphisms were selected for a multivariate ANOVA model if they showed an univariate association with TG (after adjustment for HDL-C and LDL-C) in more than 50% of bootstrap samples that were made from the original data. The multivariate model included 512 men with coronary artery disease from the REGRESS study who were completely genotyped for eight polymorphisms selected in the univariate procedure (ie, APOA1 G(-75)A, ABCA1 C(-477)T, ABCA1 G1051A, APOC3 T3206G, APOE Arg158Cys, LIPC C(-514)T, LPL Asn291Ser and LPL Ser447Stop). The gene variants APOA1 G(-75)A (P=0.04) and LPL Asn291Ser (P=0.03) were significantly associated with plasma TG levels in this multivariate analysis. The eight polymorphisms explained 8.9% of the variation in plasma TG levels. In conclusion, this study showed statistically independent effects of gene variants in the APOA1 and LPL genes on fasting plasma levels of TG. Nevertheless, only a small part of variation in TG levels could be explained by the polymorphisms.