A study was made of unoccupied androgen binding sites in the nuclei of ventral prostate glands of male rats. They were measured at 0 degrees C by comparing specific binding of 1 nM [3H]DHT to salt extract of purified nuclei during the first hour with specific binding during both hours. This method was dependent upon demonstrated completion of uptake into unoccupied binding sites within the first hour and linearity of exchange with occupied binding sites during both hours. Unoccupied binding sites were not artefactual. They did not increase if tissue concentration was diluted prior to homogenization, while they decreased if homogenization was delayed after the tissue was minced. They could be occupied, both in vitro (if precharged with at least 1 nM unlabeled DHT) or in vivo, by administering testosterone propionate subcutaneously or by infusing testosterone into the jugular vein. Exposure to a high concentration of unoccupied prostatic cytosolic binding sites (608.4 fmol from castrated rats) as compared to low concentration (29.3 fmol from intact rats) during homogenization had little effect upon nuclear unoccupied binding site concentrations (2.16 fmol/mg DNA vs 2.41 fmol/mg DNA, respectively). In individual rats, concentration of unoccupied nuclear androgen binding sites was 4.61 +/- 1.05 fmol/mg DNA, while total binding site concentration (measured with 10 nM [3H]DHT for 24h at 12 degrees C) was 866 +/- 103 fmol/mg DNA. Unoccupied nuclear binding sites reached their highest concentration in animals 4 months old (15.09 fmol/mg DNA) when animals 21 days through 720 days of age were studied. By use of association and dissociation rates of binding, it was determined that the apparent Kd of nuclear binding sites was 1.11 X 10(-12) M. There were no observed differences between unoccupied and occupied binding sites in steroid specificity or in sedimentation rate in an 8-24% glycerol density gradient. Although no physiological importance can be attributed as yet to unoccupied nuclear androgen binding sites in the prostate, they do provide a convenient comparison with putative androgen binding sites in the nuclei of testicular and epididymal germ cells.