Using samples of polygamous and non-polygamous households from villages in rural areas south of Kano, Northern Nigeria we test basic theories of household behaviour. Husbands and wives play two variants of a voluntary contributions game in which endowments are private knowledge, but contributions are public. In one variant, the common pool is split equally. In the other treatment the husband allocates the pool (and wives are forewarned of this). Most partners keep back at least half of their endowment from the common pool, but we find no evidence that polygynous households are less efficient than their monogamous counterparts. We also reject a strong form of Bergstrom’s model of polygyny in which all wives receive an equal allocation. In our case, senior wives often receive more from their husbands, no matter what their contribution. Thus the return to contributions is higher for senior wives compared to their junior counterparts. When they control the allocation, polygynous men receive a higher payoff than their monogamous counterparts. We speculate on the implications of this pattern of investment and reward for the sustainability of polygynous institutions