Colon cancer is the second most common cause of cancer-related deaths in the United States. Despite significant advances in therapy, poor prognosis of colon cancer metastasis results in high mortality. Metastasis occurs due to defects in multiple genes and consists of a series of complex sequential events that define the ultimate stage of tumour progression. Unlike tumourgenesis, which requires presence of mutated genes, metastasis is mostly a result of altered expression of genes. cDNA microarrays provide an ideal tool for the high-throughput analysis of gene function on a genome-wide scale. We assembled a collection of cDNA clones representing more than 40,000 distinct genes, developed laboratory hardware and protocols, and created databases and data analysis tools necessary for analysing differential expression. We used High-density cDNA microarrays containing more than 19,200 PCR-amplified clones to study differential expression patterns between less metastatic (KM12C) and highly metastatic (KM12L4A, KM12SM) cellular phenotypes in human colon carcinoma cell lines. Our statistical analysis of the expression ratios calculated from the measured fluorescent intensities suggests genes of prognostic or diagnostic value. Identification of these genes would assist in providing a more complete understanding of gene function and regulation with respect to cancer metastasis.