Increasing awareness of the prevalence of colorectal cancer (CRC) has resulted in a large effort to establish more informative animal-based models to elucidate the nature of this disease. One of the most widely used approaches utilizes the laboratory mouse in a variety of genetic and chemical-based models to study both inherited and sporadic (non-inherited) colorectal cancers. Colorectal cancers are classified into two morphological categories, polypoid or flat. Growing evidence suggests that flat CRCs account for 10-20% of all CRCs and that these lesions are more difficult to detect and are frequently associated with more advanced pathologies. We report using the azoxymethane (AOM) model for human CRC in combination with serial colonoscopic and histologic analyses that flat CRCs arise through a flat adenomatous intermediate rather than de novo as previously suggested. Like polypoid tumors, all flat tumors show a significant increase in nuclear beta-catenin (CATNNB1) supported by similar frequencies of mutations in the phosphorylation domain-coding region of Catnnb1. However in contrast to previous reports, tumors bearing higher "oncogenic potential" do not cluster in codon 41 of Catnnb1. Additionally, there are no differences in the frequency of mutations in codons 12 and 13 of Kras or codon 624 of Braf. Upon performing whole genome mouse microarray analyses, we found no significant differences in gene expression between flat and polypoid adenomas, however we did observe significant and mutually exclusive changes in gene expression between flat and polypoid adenomas compared to the normal colon. Based on these findings we hypothesize that the mechanism(s) which control formation of flat versus polypoid cancers lies within the normal colon and is strongly influenced by genetic background. We present here a list of candidate genes, which are differentially expressed between flat and polypoid tumors compared to the normal colon that may function in the determination of tumor morphology. Our work may provide insight into the mechanism(s) by which these distinct CRC morphologies develop and may serve as a foundation on which to identify novel genetic markers that will allow for the identification of individuals at increased risk for developing flat CRC.