The long arm of human chromosome 21 likely contains several hundred genes. To determine which of these are responsible for specific aspects of the Down Syndrome phenotype, protein functional analysis coupled to phenotypic analysis of transgenic mice will be required. Because such experiments are both time consuming and expensive, prioritizing 21q genes for further studies would be advantageous. Here, we discuss expression analysis, specifically the use of Northern analysis, cDNA array screening and RNA tissue in situ hybridization to assess place and time of expression of forty-two genes. For a subset of these, over expression in normal versus trisomy cell lines and mouse tissues is discussed. Lastly, several examples of alternative processing and their potential for generation of brain specific proteins are described. Together, these experiments give information on time, place and level of expression of a number of 21q genes and suggest some interesting candidates worth further investigation for relevance to Down Syndrome. These data also illustrate the complexities and ambiguities inherent in interpretation and use of expression information.