To date, plasma levels of acylation-stimulating protein (ASP) have been determined only in normal and obese individuals. Accordingly. ASP was measured in fasting samples obtained from 59 age-matched controls and 208 patients with documented coronary artery disease (CAD). Overall, plasma ASP was significantly higher in the CAD subjects compared to the control subjects (55.3 +/- 1.8 nmol/L CAD versus 32.0 +/- 2.6 nmol/L control, P < .0005). In the control group, the distribution of plasma ASP values was unimodal whereas in the coronary group it was significantly skewed to the right. The coronary group was subdivided into those with pronounced elevation of apoB (a marked type II phenotype, n = 13), those with hypertriglyceridemia with a normal apoB (n = 17), and the remaining CAD subjects (n = 178). In the first two groups, ASP did not differ significantly from control subjects (43 +/- 2.8 nmol/L and 49 +/- 4.4 nmol/L respectively). By contrast, in the remaining CAD subjects, both the mean ASP level (56.8 +/- 2.0 nmol/L, P < .001 by ANOVA) and the proportion of patients with a markedly elevated ASP (25.3% were > 95th percentile, P < .005 versus control by X2) were significantly increased. When this third group was divided into tertiles by plasma apoB and triglyceride there was a direct relationship between plasma ASP and these two parameters. Linear regression analysis demonstrated an association between plasma ASP and plasma triglyceride (P < .05), VLDL cholesterol (P < .025), and VLDL apoB (P < .05). Finally, when all of the CAD subjects were divided by apoE phenotype, there appeared to be a relationship between plasma ASP and apoE phenotype such that ASP was higher in E2 subjects, intermediate in E3 subjects, and lower in E4 subjects. The present data document plasma ASP levels in a number of dyslipoproteinemic states and suggest a relation between the adipsin-ASP pathway and other metabolic determinants of lipoprotein metabolism.