Surrogate biomarkers for risk assessment and efficacy of potential chemopreventive agents are needed to improve the efficiency and reduce the cost of conducting chemoprevention trials. In addition to criteria of sensitivity, specificity, quantifiability, and reproducibility applicable to most potential biomarkers, there are additional specific constraints in developing biomarkers for specific organ sites. In the case of breast tissue, these difficulties include lack of a consensus on the nature of premalignant lesions and the histologic criteria used to define them; even when such a consensus can be evolved, there are limitations in visualizing such lesions without invasive biopsies. Also, knowledge of specific genetic and biochemical changes in premalignant lesions is limited. In addition, the physiology of breast tissue is cyclic; no proven, relevant markers can be studied in a randomly obtained needle aspirate. The earliest determinate lesion that can be recognized in breast tissue is ductal carcinoma in situ (DCIS). At the University of Texas M.D. Anderson Cancer Center, we have initiated a study to develop biomarkers for tamoxifen and 4-hydroxyphenylretinamide by administering one or both of these drugs to women with DCIS or small invasive lesions in the interval between the initial diagnostic core biopsy and definitive surgery. The treatment is to be administered for 2 -4 weeks. Proposed biomarkers to be studied include: (a) markers associated with neoplastic phenotypes, e.g., excessive proliferation, alterations of nuclear morphology and angiogenesis; (b) proteins likely to be required for response to the putative chemopreventive agents, e.g., estrogen receptor, nuclear retinoid receptors; (c) markers indicative of intact downstream response pathways, e.g., progesterone receptors; (d) oncogenes and tumor suppressor genes regulated by the proposed chemopreventive agents, e.g., neu, TGF-P; and (e) potential novel markers of genetic instability that could be studied in randomly obtained needle aspirates, i.e., random chromosomal gains and losses in high risk mammary epithelium. The experience gained in designing and conducting this trial is expected to facilitate development of future chemoprevention trials of breast, as well as other organ site cancers.