After more than a decade of study, investigators are grappling for a consensus regarding the relationship between variation in candidate genes and plasma triglyceride concentration. Certain variants of LPL--both rare variants, in the case of loss-of-function mutations in kindreds with chylomicronemia, and common variants, in the case of the D9N and N291S variants--appear to be fairly consistently associated with an elevated plasma triglyceride level. In addition, the variation of the recognition site for Sstl within the 3'-untranslated region of APOC3 has consistently shown an association with a variation in plasma triglycerides. The LPL and APOC3 variants thus have at least a chance in future clinical applications, but this will require more study. Common variants of some other promising candidate genes, such as HL, have not shown as consistent an association with the variation in plasma triglyceride level. Finally, studies of variants of newer candidates, such as the mitochondrial genome, LMNA, and IL-6, indicate that many different genes might be important determinants of plasma triglyceride concentration in the general population. As always, the associations of genes with a complex intermediate trait such as plasma triglyceride level depend upon interactions with modulatory factors such as genetic background and/or secondary genetic effects, in addition to the effects of gender, age, hormone replacement, and postprandial status. A key attribute for increasing confidence in the biologic or potential clinical validity of the associations of candidate gene variation with plasma triglyceride will be the development of assays that will provide a more direct mechanistic link between the genetic variant and the elevated plasma triglyceride.