To define the molecular changes associated with ovarian cancer, DNA microarray analysis has been adapted to detect differentially expressed genes in human normal ovary tissue, borderline, and invasive epithelial ovarian tumors. The differential expression of genes in the tumor tissues and normal tissues was confirmed by Northern and/or semi-quantitative RT-PCR analysis. Analysis of the differential gene-expression profiles of the normal and neoplastic ovary allowed us to detect previously unidentified genes in ovarian tissues. We observed up-regulation of the following genes in ovarian cancer: catechol-O-methyltransferase (COMT), the autocrine motility factor neuroleukin (NLK), the transcription regulator high mobility group I proteins (HMGI), the tyrosine kinase receptor ErbB-3, S100-alpha protein and Acyl-CoA-binding protein (ACBP). The transcription factor, chicken ovalbumin up-stream promoter transcription factor II (COUP-TFII), was the only gene down-regulated in ovarian cancer. Comparable gene-expression profiles were previously reported in breast cancer, suggesting that similar molecular events also exist in ovarian cancer. Our microarray analysis showed that most differentially expressed genes in ovarian cancer are linked to glucose/insulin metabolism, providing a possible molecular link between the glucose/insulin signaling pathway and the neoplasms of ovarian cancer.