Our previous results had shown that brain cytosol from androgen-insensitive, testicular feminized (Tfm/y) mice contains a reduced amount (20-25%) of androgen receptors, compared to normal female or castrated male mice, with unchanged affinity for dihydrotestosterone (DHT). We have now used various physicochemical techniques to ask whether there is a qualitative difference between these residual receptors and those in normal brain. When androgen receptors labeled with [3H]DHT in concentrated crude brain cytosol were analyzed by density gradient centrifugation and Agarose gel filtration in buffers containing 0.4-0.5 M KCl, Tfm/y receptors appeared to be smaller and more symmetrical than those from their normal siblings; however, if wild-type receptors were partially purified or prepared from slightly more dilute homogenates, their properties approached those of Tfm/y receptors in crude cytosol. This suggested that receptor molecules in concentrated cytosol from normal mice were aggregated to a greater extent than were those from the mutant. The molecular weight (54,000) and axial ratio (3:1 for prolate or oblate ellipsoid) calculated for Tfm/y receptors, therefore, may provide estimates for normal receptors as well. By DNA-cellulose chromatography, Tfm/y and female cytosol receptors were both resolved into two components, eluting at about 0.13-0.15 and 0.22-0.24 M NaCl. On DEAE-cellulose columns, both were eluted as a single major species at approximately 0.08-0.10 M KCl. Thus, excluding their state of aggregation, presumably resulting from their different concentrations in crude cytosol, Tfm/y and normal receptors were substantially identical as concerns the physical parameters examined in this study.