One of the challenges to the effective use of cDNA array analysis in mouse models of oncogenesis is the choice of a critical set of probes which are informative for human disease. Given the thousands of genes with a potential role in human oncogenesis and the hundreds of thousands of mouse sequences available for use as probes, assimilation of an overlapping set can be an overwhelming task. We have developed a bioinformatics tool that will mine a source of mouse sequences. It presents the output in a form that enables the user to access the information necessary to define a list of sequences that will enhance gene discovery and functional assessment in mouse models of oncogenesis. This web-based sequence mining tool lists publicly available sequences in order of importance to oncogenesis in humans and provides a focused set of informative links including information on clone source and, if known, gene function. The Mouse Oncochip Design Tool uses the Mouse Genome Database (MGD) developed and maintained by the Jackson Laboratories for mouse DNA sequences. There are over 380,000 sequences in this database. The user can choose a subset of sequences based on expression in a specified tissue source, chosen from over 300 types of tissues. The output list has been ordered to present the most likely candidates first, as defined by identification of a mouse gene associated with the sequence, identification of a human homologue for the mouse gene marker and inclusion of the human homologue in an oncogene set. The tool is web based and uses a Perl common gateway interface (CGI) to the browser and Standard Query Language (SQL) scripts embedded in Data Base Independent Perl (DBI Perl) to extract the necessary information from the Mouse Genome Database and the Human Oncogene set (which was compiled in the Division of Clinical Sciences at the National Cancer Institute). Links to the clone source and annotation for the mouse genes stored in the MGD, to GenBank and the Unigene cluster resource of the National Center for Biotechnology Information (NCBI), and to GeneCards (Weizmann Institute) are constructed dynamically by the code so that there is no need for updating. This resource is part of a bioinformatics toolkit for studies of mouse models of human cancer which uses microarray technology.