Single nucleotide polymorphisms (SNPs), the most common form of human genetic variation, have been touted as useful markers for dissecting the genetic basis of complex diseases and predicting differences in drug response. It is hoped that disease profiling and chemopredictive testing will one day enable patients to be screened for disease and guide the therapeutic course of action—so-called pharmacogenomics. SNP discovery has now begun in earnest, with the US National Institutes of Health (Rockville, MD) starting work on a genome wide map of 100,000 markers and a consortium of 10 pharmaceutical companies and the Wellcome Trust planning to generate 300,000 publicly available SNPs within 2 years. While these efforts promise to rapidly collate SNP data, the feasibility of extensive SNP-based analysis of complex disease and drug response remains widely debated. On p. 505, McCarthy and Hilfiker discuss the practical issues in realizing the potential of SNPs for dissecting drug response and disease traits. Their review describes how patient sample size, SNP density and genome coverage, statistical significance, and data interpretation will be important in designing large-scale studies for the reliable identification fof genetic associations.