Abstract
A synopsis is presented of hormone-cancer relationships as seen by an epidemiologist. Emphasis has been placed on reproductive risk factors and cancers of the endometrium, ovary and breast. Although there are definite parallelisms in this regard the aetiological mechanisms for these three sites may be different. Evidence is accumulating that the protective effect of an early first birth and of subsequent pregnancies on breast cancer risk can be understood biologically in terms of interference with the occurrence of preneoplastic lesions.