We integrate individual power in groups into general equilibrium models. The relationship between group formation, resource allocation, and the power of specific individuals or particular sociological groups is investigated. We introduce, via an illustrative example, three appealing concepts of power and show that there is no monotonic relationship between these concepts. Then we examine existence of competitive equilibria with free exit and study whether maximal individual power is consistent with Pareto efficiency. As applications, we discuss when power spillovers occur and we identify human relation paradoxes: positive externalities increase, but none of the household members gains in equilibrium. We further identify implicit, determinate and de facto power.