Under the auspices of the Nuclear Receptor Signaling Atlas (NURSA), we have undertaken to evaluate the feasibility of targeting nuclear receptor-coactivator surfaces for new drug discovery. The underlying objective of this approach is to provide the research community with reagents that can be used to modulate the transcriptional activity of nuclear receptors. Using combinatorial peptide phage display, we have been able to develop peptide antagonists that target specific nuclear receptor (NR)-coactivator binding surfaces. It can be appreciated that reagents of this nature will be of use in the study of orphan nuclear receptors for whom classical ligands have not yet been identified. In addition, because the interaction of coactivators with the receptor is an obligate step for NR transcriptional activity, it is anticipated that peptides that block these interactions will enable the definition of the biological and pharmacological significance of individual NR-coactivator interactions. In this report, we describe the use of this approach to develop antagonists of the androgen receptor by targeting its coactivator-binding pocket and their use to study the coactivator-binding surface of this receptor. Based on our findings, we believe that molecules that function by disrupting the androgen receptor-cofactor interactions will have use in the treatment of prostate cancer.