We revisit and test Salop and Stiglitz (1982) Theory of Sales. Equilibrium comparative static predictions are that greater consumer storage constraints lead to: (1) higher average prices, (2) fewer promotions, and (3) shallower promotions. In equilibrium, price dispersion is nonlinear in storage constraints, first increasing then decreasing. Empirical estimates of storage constraints are developed for approximately 1,000 households using the American Housing Survey (1989), United States Census (1990), and Stanford Market Basket Database (1991–1993). We find consumers with greater storage constraints shop more often and purchase smaller quantities per visit; moreover, the comparative static predictions are supported and evidence consistent with the equilibrium dispersion prediction is observed. Estimated quantitative effects are economically important.