Besides the classical land rent theory, a few theoretical hypotheses exist in the explanation of agricultural land use changes in terms of both areal extent and intensity. Any change in land use has its impact on environment and ecology. And these land use hypotheses are usually extended to general hypotheses about human-land relations while environmental impacts of land use changes are taken into account. This paper reviews theoretical hypotheses about agricultural land use changes. These include ①compelled areal expansion hypothesis based on Malthus′s demographic theory; ②Boserup′s induced intensification hypothesis; ③Geetz′s elastic intensity hypothesis and ④Brookfield′s disintensification hypothesis. Some relevant propositions about environmental impacts of agricultural land use changes are also discussed. They are: ①farmland expansion results in land degradation; ②agricultural intensification of fertile lands indirectly helps improve ecology in fragile environments and ③environment and ecology benefit from agricultural extensification and farmland abandonment. Empirical evidences from both domestic and foreign sources related to the four hypotheses and three propositions are reviewed and discussed. It is concluded that application of the four hypotheses and three propositions face severe limitations in the explanation of agricultural land use changes and their environmental impacts in different context. These limitations, however, cannot offset their merits as bases for vision building and scenario analysis in the process of land use planning.