Androgen receptor (AR) antagonists are important compounds for the treatment of prostate cancer (PCa). The atraric acid (AA), a natural compound, binds to the AR and acts as a specific AR antagonist. Interestingly, AA represents a novel chemical platform that could serve as a potential basis for new AR antagonists. Therefore, one objective of this study was to analyze the chemical/structural requirements for AR antagonism and to obtain predictions of where and how AA binds to the AR. Further, this study describes the chemical synthesis of 12 AA derivatives and their analysis using a combination of computational and functional assays. Functional analysis of AA derivatives indicated that none activated the AR. Both the para-hydroxyl group and the benzene ortho- and the meta-methyl groups of AA appeared to be essential to antagonize androgen-activated AR activity. Furthermore, extension of the hydrophobic side chain of AA led to slightly stronger AR antagonism. In silico data suggest that modifications to the basic AA structure change the hydrogen-bonding network with the AR ligand binding domain (LBD), so that the para-hydroxyl group of AA forms a hydrogen bond with the LBD, confirming the functional importance of this group for AR antagonism. Moreover, in silico modeling also suggested that the ortho- and meta- methyl groups of AA interact with hydrophobic residues of the ligand pocket of AR, which might explain their functional importance for antagonism. Thus, these studies identify the chemical groups of AA that play key roles in allowing the AA-based chemical platform to act as an AR antagonist.