Nuclear Hormone Receptors (NR) represent one of the most promising protein families in terms of therapeutic applications. These transcription factors are naturally switched on and off by small molecule hormones presenting physico-chemical properties very similar to therapeutic chemical entities. NRs represent therefore intrinsically a very good family of protein targets for the prevention and treatment of diverse diseases, including cancer. Several known anti-cancer drugs, such as tamoxifen or flutamide, are targeting NRs, and many more are expected to reach market. The detailed knowledge of the structural mechanism underlying activation and inhibition of NRs by small molecule modulators begets important therapeutic opportunities. The crystal structure of at least nine NR ligand binding domains (LBDs) revealed at the atomic level how natural or synthetic agonists and antagonists can promote recruitment of co-activator and co-repressor proteins. Interestingly, it was recently shown that nucleotide polymorphisms located in NR LBDs could alter or even reverse the response of the receptors to small molecule ligands. Mapping these polymorphisms on the structure of the LBD can reveal why agonists or antagonists become inactive against the mutated receptor, allow atomic models for resistance to cancer therapy, and open the door to the rational design of improved anti-cancer drugs, customized for each patient.