Microarray-based genomic techniques allow the simultaneous determination of relative levels of expression of a large number of genes. Studies of the transcriptome in complex neurobiological systems are uniquely demanding due to the heterogeneous nature of these cells. Most brain regions contain a large variety of cell populations that are closely intermingled. The expression of any specific gene may be restricted to a subpopulation of cells, and changes in gene expression may occur in only a small fraction of the cells expressing that transcript. Due to this dilution effect, many genes of interest are expected to have relatively low levels of expression in tissue homogenates. Furthermore, biologically significant differences in expression may result in only small fold-changes. Therefore genomic approaches using brain dissections must be optimized to identify potentially regulated transcripts and differential expression should be confirmed using quantitative assays. We evaluated the effects of increasing tissue complexity on detection of regulated transcripts in focused microarray studies using a mouse cell line, mouse hypothalamus and mouse cortex. Regulated transcripts were confirmed by quantitative real-time PCR. As tissue complexity increased, distinguishing significantly regulated genes from background variation became increasingly more difficult. However, we found that cDNA microarray studies using regional brain dissections and appropriate numbers of replicates could identify genes showing less than 2-fold regulation and that most regulated genes identified fell within this range.