There is currently great interest in identifying genetic loci associated with the expression of one or more genes by combining methods for the mapping of quantitative trait loci (QTL) with microarray technology. Because of the high cost of microarrays, it is useful to develop strategies for selecting a subset of individuals for expression profiling. We describe a method of selective transcriptional profiling that uses available information on individual quantitative traits and marker data to choose an optimal set of individuals for measurement with microarrays. We show how to analyse the resulting data to identify genes whose expression is associated with a QTL for a traditional trait or any other locus. Examples are used to illustrate the method and highlight some challenges associated with the analysis of selected data.