It is 10 years since the discovery of the human genome; however, the study of the influence of genetic variants on drug effect - pharmacogenomics - has so far failed to create a major impact on day-to-day prescription practices. In the present article we analyze the main findings in candidate gene variants, gene combinations and whole-genome scans in relation to diuretic treatment. A critical analysis of the main reasons for some contrasting results will be discussed. The hypertension phases, in clinical trials dealing with genes and related pathophysiological mechanisms, may account for these inconsistent findings. The use of previously untreated versus treated patients is addressed. Finally, a positive study with a new genetic molecular strategy is described.