Cancers of the same primary site are in many cases heterogeneous in molecular pathogenesis, clinical course, and treatment responsiveness. Current approaches for treatment development, evaluation, and use result in treatment of many patients with ineffective drugs and lead to the conduct of large clinical trials to identify small, average treatment benefi ts for heterogeneous groups of patients. New genomic and proteomic technologies provide powerful tools for the identifi cation of patients who require systemic or aggressive treatment and the selection of those likely or unlikely to benefi t from a specifi c regimen. In spite of the large literature on developing prognostic and predictive biomarkers and on statistical methodology for analysis of high dimensional data, there is considerable uncertainty about proper approaches for the validation of biomarker-based diagnostic tests. This article attempts to clarify these issues and provide a guide to recent publications on the design of clinical trials for evaluating the clinical utility and robustness of prognostic and predictive biomarkers.