Many predict that pharmacogenomics is poised to deliver on the promises of the genomic revolution in ushering an era of personalized medicine. However, questions have emerged over whether the field will deliver a truly individualized medicine or if population-based therapies that build on conventional notions of racial biology will prevail. At the heart of this issue is the challenge of knowing which axes of stratification are appropriate in identifying population differences and to what extent is race and/or ethnicity an appropriate method of comparison in studies of genetic variation. These questions make plain that in addition to the development of technical tools to identify salient gene variants associated with drug response, serious consideration over how best to characterize populations in human genetic variation research must be given in order to realize the putative benefits of tailored therapeutics.