The opportunity costs of rearing British children in terms of cash earnings foregone by their mother are estimated for a typical family. Data from the Women and Employment Survey (1980) provide estimates for hourly pay as a function of work experience and current hours of work. They also make possible the generation of participation and hours profiles for representative women with different numbers of children, which, together with the earnings function are used to simulate lifetime income. Earnings foregone as a result of bearing and rearing two children are roughly equally composed of three effects, on participation, hours and pay. They exceed direct costs, and do not rise proportionally with family size. They are sensitive to the spacing of births but not, undiscounted, to the timing of the first one. The method and results contrast markedly with those of a similarly motivated study of US women, by Calhoun and Espenshade. It is argued that the non-linearity of the earnings function and state dependence in British labour force transitions would violate the assumptions which permitted the US exercise to be based on a multi-state labour force life-table.