The Farr assay is thought to detect only antibodies to DNA of relative high avidity. This is due to the high salt concentration of the employed ammonium sulfate precipitation, which dissociates DNA-anti-DNA complexes of low avidity. A recently introduced method to detect anti-DNA, the PEG assay, circumvents these dissociating reaction conditions by using polyethylene glycol instead of ammonium sulfate to precipitate the complexes; we therefore thought to measure antibodies to DNA of low avidity as well. We tested this assumption in several ways. It was found that the PEG assay detects a population of antibodies to DNA that are missed by the Farr assay. Complexes made with these antibodies were salt labile and could readily be dissociated by means of excess DNA, whereas Farr-positive antibodies formed stable complexes with DNA. Avidity studies using the method described by Celada et al. indicated that the anti-DNA detected by the PEG assay but missed by the Farr assay was of relatively low avidity. An inverse correlation between avidity and slope of the binding curves in the PEG assay was observed. These results confirm the notion that the PEG assay detects antibodies to DNA of low avidity. The fact that the Farr assay does not measure these antibodies confers possible diagnostic importance upon the PEG assay.