Estrogens are potent mitogens for a subset of their target tissues, including those in breast cancer, in which they promote growth by interacting with specific intracellular receptors (ER-α and -β) of the nuclear receptor superfamily of transcription factors. Definition of the gene expression patterns, or signatures, characteristic of estrogen in growth-responsive cells would allow identification of the gene networks that mediate the cellular responses to these hormones and the discovery of new hormone-responsive genes and gene products that could be used as markers in experimental and clinical settings. For this purpose, we performed an initial characterization of the gene responses of human breast cancer cells to estrogen by determining the gene expression signatures of these cells before and after hormonal stimulation. We used microarrays containing more than 9,000 unique gene probes, about half of them unnamed. About 40% of the genes analyzed, including close to 1,200 expressed sequence tags, were detectable in RNA extracted from human breast cancer cells. Furthermore, the concentration of 10% messenger RNAs was either increased or decreased in the cell at one time or another after hormonal stimulation, each such change corresponding to a specific cell cycle phase. These included transcripts from known estrogen-responsive genes (such as pS2, c-fos, c-myc, and cyclin D1) as well as a significant number of new hormonal targets. Hierarchical clustering shows distinct patterns of gene activation and inhibition, which will be exploited to further investigate intracellular signaling by steroid hormones in human breast cancer cells.