Owing to its clinical accessibility, peripheral blood is probably the best source for the assessment of differences or changes in gene expression associated with disease or drug response and therapy. Gene expression patterns in peripheral blood cells greatly depend on temporal and interindividual variations. However, technical aspects of blood sampling, isolation of cellular components, RNA isolation techniques and clinical aspects such as time to analysis and temperature during processing have been suggested to affect gene expression patterns. We therefore assessed gene expression patterns in peripheral blood from 29 healthy individuals by using Affymetrix microarrays. When RNA isolation was delayed for 20-24 h-a typical situation in clinical studies-gene signatures related to hypoxia were observed, and downregulation of genes associated with metabolism, cell cycle or apoptosis became dominant preventing the assessment of gene signatures of interindividual variation. Similarly, gene expression patterns were strongly dependent on choice of cell and RNA isolation and preparation techniques. We conclude that for large clinical studies, it is crucial to reduce maximally the time to RNA isolation. Furthermore, prior to study initiation, the cell type of interest should already be defined. Our data therefore will help to optimize clinical studies applying gene expression analysis of peripheral blood to exploit drug responses and to better understand changes associated with disease.