Conjugates of testosterone-3-carboxymethyloxime (T-3-CMO), testosterone-17-hemisuccinate (T-17-HS), 17 beta-estradiol-6-carboxymethyloxime (E-6-CMO), or 17 beta-estradiol-17-hemisuccinate (E-17-HS) and bovine serum albumin (BSA) with varying steroid:protein ratios were prepared using the mixed anhydride method. Dialysis followed by molecular filtration yielded monomer steroid-BSA conjugates with a molecular weight of 70,000 dalton, and polymer conjugates with molecular weights of 140,000 dalton and higher. When conjugates were prepared with increasing initial steroid:BSA molar ratios the ratio of the obtained conjugates increased, in parallel with a decrease in the relative amount of monomers and an increase in the mean molecular size of polymers. The molecular properties of these conjugates were studied further by polyacrylamide gel electrophoresis (PAGE) in native and denaturing conditions. In native PAGE the monomer fractions showed one main band with a mobility slightly lower than BSA and a faint band corresponding with BSA-dimers. The polymer fractions consisted of a heterogeneous population of protein oligomers with molecular weights varying from 140,000 to over a million dalton. In the presence of sodium dodecylsulphate part of the polymers dissociated into monomers. In buffered aqueous solutions the bulk of the conjugate preparation retained its molecular size and composition, although the generated covalent bonds were found to be liable to spontaneous hydrolysis. Steroid-protein conjugates were shown to contain appreciable amounts of non protein-bound steroids. Binding of T-BSA to androgen receptors in rat ventral prostate cytosol was assayed using LH-20 chromatography and sucrose gradient centrifugation analysis. Binding of E-BSA to estrogen receptors was analysed with rat uterus cytosol using the dextran coated charcoal assay and the sucrose gradient centrifugation technique. Relative binding affinities (RBA) were analyzed in competition experiments using radiolabeled ligands. It was found that the molecular size of the conjugate does not influence its interaction with steroid receptors. Steroid coupled via the 17-position show a higher RBA to receptors than the T-3 or E-6 derivatives. The RBA of T-3-BSA, T-3-CMO, T-17-BSA and T-17-HS appeared to be very low, i.e. between 0.1 and 1.7% of the RBA of dihydrotestosterone. Consequently, high concentrations of conjugate are required to saturate androgen receptor binding sites. Under these conditions involvement of type II and eventually type III binding sites, which show less ligand specificity and lower affinity, may be anticipated preventing exclusive detection of androgen receptors.(ABSTRACT TRUNCATED AT 400 WORDS)