The subject of this article is the relationships between farm size and prod
uctivity and between farm size and profitability in the developing countrie
s. The recent controversies over the inverse size-output relationship are r
eviewed, and a framework is provided that explains the inverse relationship
based on plausible assumptions about imperfections in the markets for labo
r, land, credit and risk. From this framework, a set of testable hypotheses
are derived. The hypotheses are tested on recent farm-level panel data fro
m Pakistan. A strong inverse relationship between farm size and yield is pr
esent in the sample, even when household fixed effects are used to account
for unobserved heterogeneity, Moreover, the suggested market imperfections
framework is consistent with the data. (C) 1998 Elsevier Science Ltd. All r
ights reserved.