This paper reviews and analyzes a literature that examines the restrictiveness of focusing on households as single units of analysis. In particular, this literature asks: To what extent can we use the household as a unit of analysis, and when do we need to disaggregate the household and look at individual behavior within the household? In this paper, I examine the different models of intrahousehold resource allocation and the empirical work they have generated. Each of the models has a different set of assumptions about the four components of household decisions. The assumptions and predictions of each of the main models of intrahousehold resource allocation are presented; then I discuss the empirical work that tests the validity of the assumptions and discuss when the different models may be appropriate.