Animal societies vary markedly in reproductive skew, the extent to which breeding is monopolised by dominant individuals. In the last few years, a large number of different models have been developed to explain this variation. Here, I review existing models of reproductive skew, distinguishing between two basic types. Transactional models focus on group stability and the constraints this places on the division of reproduction. Compromise models, by contrast, ignore issues of group stability and view the division of reproduction as the outcome of a conflict in which each group member has a limited or partial ability to enforce its own optimum. I go on to show, however, that the division between transactional and compromise models is somewhat artificial, and that both approaches may be combined in a single, synthetic treatment. Different models of reproductive skew are thus better seen as special cases of a general underlying theory, rather than alternative paradigms. I conclude with a brief discussion of the possibilities and problems of empirically testing this unified theory of skew, and the prospects for future theoretical advances.