It has been suggested that oestriol protects against breast cancer, because in some experiments on uterine growth it is only weakly active, and partially inhibits the effects of oestradiol-17beta. When its effects are measured 24 h after a single injection, oestriol behaves as a typical impeded oestrogen with low potency and a flat dose-response line. This does not result from failure to stimulate certain critical stages of growth but from failure to sustain the products of growth. We found that oestriol induced all phases of uterine growth including DNA synthesis and cell division. It was as effective as oestradiol in stimulating early increases in protein synthesis and uterine weight, and half as effective in stimulating epithelial cells to replicate DNA and divide. However, epithelial cell numbers did not increase after a single injection of oestriol because cell death rate increased at the same time as mitotic rate, apparently as a result of the more rapid loss of oestriol from the uterus. Repeated injections of oestriol prevented premature cell death and produced as much uterine hypertrophy and hyperplasia as oestradiol-17beta. These results support the thesis that the oestrogenic potency of a substance is largely determined by the duration of its occupation of receptors. Thus in situations of continuous production, (e.g. pregnancy) oestriol would be as active as oestradiol and unlikely to exert any significant 'buffering' or protective action. The findings are also discussed in relation to a new model for the regulation of cell proliferation.