Cervical cancer chemoprevention agents under study include diet and micronutrients (particularly beta-carotene, folate, and vitamins A, C, and E); medications such as retinoids (retinyl acetate gel, all-trans-retinoic acid, and 4-hydroxyphenylretinamide) that are chemically related to micronutrients; and other chemopreventives meant to affect the carcinogenic process at the cellular level, including such polyamine synthesis inhibitors as alpha-difluoromethylornithine. Agents become reasonable candidates for study when they have a biologic rationale, they are of low toxicity, and they can be taken for a long period of time. Since the human papillomavirus (HPV) is the major etiologic agent, the medication should show activity against HPV-positive preinvasive and invasive cell lines. The medication needs to be of low toxicity because it may be taken for long periods of time and less toxicity is tolerated in the precancerous setting. Until 1995, none of the studies used surrogate end point biomarkers (SEBs), relying instead on histologic and colposcopic regression as end points. All studies typically included subjects with cervical intraepithelial neoplasia. Conclusions to be drawn from these studies include the following: Though micronutrients are logical candidates for chemoprevention, they haven't worked consistently, and the reasons remain unclear. Furthermore, SEBs need to be validated in phase I trials. Finally, a better understanding of the role of HPV needs elucidation, including an understanding of the relationship of the medication to HPV status and of the immunobiology of HPV throughout the trial.