The subject of this article is the alleged inverse relationship between farm size and productivity in developing countries. The recent controversy is reviewed, and a framework is provided to explain the inverse relationship based on plausible assumptions about imperfections in the markets for labor, credit and land. On this basis testable hypotheses are derived. Using fram.level panel data from Pakistan, the framework is assessed by regressing output on operational fram size, size of owned holding, family size, tenurial status and irrigation status of the land. Household fixed effects are used to account for remaining unobserved heterogenity. It is concluded that an inverse relationship is present in Pakistan, and that the market imperfections framework performs well with the data