The challenge of genome medicine is not to make a significant move into the clinic over the next five years. The floodgates of genomic research have been opened, and the hopes are that the rising tide will spill over into medical practice in the form of diagnostic tests, risk assessment tools and therapeutics. The ability to perform genome-wide analyses using several different approaches has already provided tantalizing new clues to disease causation and therapeutic targets. Companies have sprung up to use these new technologies to provide information to individuals about predicted health and disease, and about behavioral traits. As exciting as these prospects are, it is far too soon to promise clinically useful information from genomic analyses. We have far to go to assure basic levels of analytical validity or clinical validity of the diagnostic or predictive tools on offer, and to determine their clinical utility in the medical context.