We propose to continue Case-Control Surveillance of medications and serious illnesses (particularly cancers), a system which has been in operation since 1975. The usefulness of this system for quantifying the magnitude of effects of medications on the risk of serious illnesses, or documenting safety, has been established, as has been its capability for identifying unsuspected associations. In a network of participating hospitals, nurse-interviewers administer standard questionnaires to patients admitted for recently diagnosed illnesses of special interest. Histories of medication use and information on covariates that may confound or modify the exposure-disease associations are obtained. The emphasis is on regular and long-term use of medications, particularly on those taken by relatively healthy individuals (e.g., oral contraceptives). Should a hypothesis of interest arise, the resources of the system are directed toward rapid accrual of appropriate cases if a sufficient sample has not already been collected, and modifications to the questionnaire may also be made. The accrued data are a resource that has been used for numerous in-depth analyses of hypotheses that arose from the data base itself or, more usually, from other sources. The data have also been used to document safety, with narrow confidence intervals. The data are also "screened" periodically for unsuspected associations between exposures and the several diseases. Associations observed elsewhere have generally been observed in the screen, and new screen associations have generally been confirmed in other data. Previous data collection and analyses have focused on cancers (breast, ovary, endometrium) and nonmalignant illnesses in women, mainly in relation to contraceptive and hormone use. Future data collection and analyses will concern a variety of drugs and exposures both in men and women, mainly in relation to cancers: among others, these will include colorectal cancer, lung cancer, trophoblastic disease, pancreatic cancer, cancers of the liver and testis, and lymphomas. Should an important new hypothesis arise, the resources of the system will be directed toward its evaluation.