Human cancers are diverse in their pathology and responsiveness to clinical treatment. This diversity is at least in part due to variations in cellular gene expression programs. Although the analyis of proteins--the key players in cells and potential drug targets--is advancing rapidly, there are situations in which the analysis of RNA rather than proteins can provide valuable information for the diagnosis of cancer. These situations include absense of an antibody for the protein of interest, expression of functionally defective proteins, expressed small nucleotide polymorphisms (SNPs), analysis of alternatively or abnormally spliced molecules, and functional analysis of splice site mutations. In this chapter we will focus on the analysis of RNA from clinical samples and will summarize how gene expression studies on the RNA level using a variety of new tools can be useful for discovering new classes of tumors, for predicting clinical outcome or therapy response, and for designing novel personalized clinical interventions that can not be achieved with histology alone.