Paraxonase, an enzyme associated with the high density lipoprotein (HDL) particle, hydrolyzes paraoxon, the active metabolite of the insecticide parathion. Several studies have shown that paraxonase levels in humans have a distribution characteristic of two alleles, one with low activity and the other with high activity. Paraoxonase also has arylesterase activity, which does not exhibit activity polymorphism and can therefore serve as an estimate of enzyme protein. Although the ability of paraoxon to irreversibly inhibit lipoprotein lipase (LPL) has been exploited experimentally for many years, the role of plasma paraoxonase in lipoprotein metabolism is unknown. Seventy-two normal individuals were examined for paraoxonase genotypes, plasma paraoxonase and arylesterase activities, postheparin LPL and hepatic lipase (HL) activities, and lipoprotein levels to determine whether (1) paraoxonase activity or genotype determines lipoprotein levels via an effect on LPL or HL activity or (2) variation in LPL and HL activities determines HDL levels and indirectly affects paraoxonase activity and protein levels in plasma. In the entire group, paraoxonase activity was related to arylesterase activity and genotype. Whereas arylesterase activity was correlated with HDL cholesterol (HDL-C) and apolipoproteinA-I (apoA-I) levels, neither arylesterase nor paraoxonase was correlated with LPL or HL activity. Furthermore, LPL activity was positively correlated and HL inversely correlated with HDL cholesterol and apoA-I levels, whereas LPL was inversely correlated with triglyceride levels. The paraoxonase genotypes of the study group were 30 individuals homozygous for the low-activity allele, 38 heterozygotes, and 4 individuals homozygous for the high-activity allele. Paraoxonase genotype accounted for approximately .75 of the variation in paraoxonase activity. Paraoxonase activity was linearly related to arylesterase activity within each subgroup. No difference in either LPL or HL activity was seen as a function of paraoxonase genotype, nor were differences seen in plasma triglyceride or HDL-C by genotype by ANOVA. The relation between LPL and HL and components of HDL in the paraoxonase genotypic subgroups in general reflected the associations seen in the group as a whole. Multivariate analysis showed that LPL, HL, and arylesterase, a measure of paraoxonase mass, were independent predictors of HDL cholesterol, while paraoxonase genotype or activity was not. Thus, variation in LPL and HL appears to be significantly related to HDL cholesterol and apoA-I levels. The levels of HDL are a major correlate of paraoxonase protein levels, while paraoxonase genotype is the major predictor of plasma paraoxonase activity.