Recent studies showed lower apolipoprotein A-IV (apoA-IV) plasma concentrations in patients with coronary artery disease (CAD). The actual distribution of the antiatherogenic apoA-IV in human plasma, however, is discussed controversially and it was never investigated in CAD patients. We therefore developed a gentle technique to separate the various apoA-IV-containing plasma fractions. Using a combination of precipitation of all lipoproteins with 40% phosphotungstic acid and 4 M MgCl2, as well as immunoprecipitation of all apoA-I-containing particles with an anti-apoA-I antibody, we obtained three fractions of apoA-IV: lipid-free apoA-IV (about 4% of total apoA-IV), apoA-IV associated with apoA-I (LpA-I:A-IV, 12%), and apoA-I-unbound but lipoprotein-containing apoA-IV (LpA-IV, 84%). We compared these three apoA-IV fractions between 52 patients with a history of CAD and 52 age- and sex-matched healthy controls. Patients had significantly lower apoA-IV levels when compared to controls (10.28 +/- 3.67 mg/dl vs. 11.85 +/- 2.82 mg/dl, P = 0.029), but no major differences for the three plasma apoA-IV fractions. We conclude that our gentle separation method reveals a different distribution of apoA-IV than in many earlier studies. No major differences exist in the apoA-IV plasma distribution pattern between CAD patients and controls. Therefore, the antiatherogenic effect of apoA-IV has to be explained by other functional properties of apoA-IV (e.g., the antioxidative characteristics).