A case-control study of the relationship between oral contraceptive (OC) use and cancer of the breast or cervix in women aged 45 and under is currently underway in 4 cancer centers, a public assistance hospital, and 3 private clinics in France. The preliminary analysis is presented of 240 cases of breast cancer, 84 of cervical cancer, 68 controls with nongynecological cancers, and 246 controls with nonmalignant conditions except benign breast disease and moderate or severe cervical dysplasia. 46% of controls had benign and 7% had malignant thyroid conditions, 4% had ovarian cysts, 3% had uterine fibromas, 6% had other gynecological conditions, 9% had other types of conditions, and 10% were hospital personnel, visitors, or women having routine examinations. The mean duration of OC use was not significantly different among different groups of controls. 3 factors were found to raise the risk of breast cancer significantly in the study population: a family history of breast cancer raised relative risk to 1.75, having 3 or fewer children raised it to 2.16, and having 2 or more induced abortions raised it to 1.75. Occupational status, marital status, age at menarche, age at 1st birth, number of spontaneous abortions, and benign breast disease were not associated with breast cancer. There was no association between breast cancer and duration of OC use or OC use before or after the 1st pregnancy or age 25. Women beginning OC use before age 25 had a relative risk of .55 compared to other women. No increased risk of breast cancer was seen when various types of OCs, including combined, biphasic, sequential, or low or higher dosed progestagen only formulations were analyzed separetely, and users of progestagen only pills had a decreased incidence. There was no relationship between risk of brest cancer and estrogen dose of combined, sequential, or biphasic pills. 4 factors were related to cervical cancer: high occupational status, number of children, number of induced abortions, and history of benign breast disease. OC use was not significantly related to cervical cancer. The relative risk of cervical cancer was less than unity through 5 years of use and increased to 1.39 after 9 years, but still failed to attain statistical significance. The risk of cervical cancer was found to increase with use of OCs for more than 7 years beginning after age 25. The sample size was too small to permit separate analysis for different types of OCs.