Published retrospective studies of endometrial and ovarian epithelial cancer all suggest a protective effect of combined oral contraceptives (OCs). The effect appears to increase with duration of use and to persist for several years after cessation of treatment, a significant finding since both cancers are rare in women under 40. Available data from 2 British prospective studies are sparse but also point to a protective effect. The effects of sequential OCs on ovarian cancer are unknown. Attempts to establish a relationship between cervical cancer and OC use have encountered many difficulties, among them problems in controlling for sexual behavior and other confounding variables, the fact that barrier methods protect against cervical cancer, possible distortion of results due to substantial differences in the smears of groups being compared, and possible bias introduced by disproportionate quantity of material from women using a particular contraceptive method. Numerous studies based on data from routine cervical cytological examinations have given varying results of little value. Most retrospective studies found no relationship, but studies by Ory and Harris and their collaborators found positive associations between length of OC use and risk of cervical cancer. Prospective studies have tended to give positive results. The Walnut Creek Study found a significant positive relationship between cervical cancer and duration of OC use, but later analysis showed that the majority of effects could have been caused by failing to control for sexual behavior. The most recent data, from the Oxford Family Planning Association study, compared 6838 OC users to 3154 IUD users. All participants had children, and the 2 groups were reasonably comparable in regard to the known risk factors for cervical neoplasia and to sexual behavior. Data concerning invasive cancer, cancer in situ, and dysplasia indicated a significant relation between the risk of cancer and the duration of OC use. No such relationship was found among IUD users. A large World Health Organization multicenter study of 726 cases and more than 5000 controls which included information on the participants' sexual behavior was reported to have indicated a moderately increased risk of cervical cancer paralleling increased duration of OC use. The full report has not yet appeared.