In this paper, we test the hypothesis that land held under varying configurations of property rights will be farmed at different levels of production efficiency. Production data were collected from 477 plots in a fairly productive, mixed farming system in the Ethiopian highlands. Interspatial measures of total factor productivity, based on the Divisia index, were used to measure the relative production efficiency of three informal and less secure land contracts (rented, share-cropped and borrowed) relative to lands held under formal contract with the Ethiopian government. Although the informally-contracted lands are farmed 10-16% less efficiently, the analysis indicates that farmers of such lands actually apply inputs more, rather than less, intensively (i.e., more inputs per unit of land). The gap in total factor productivity thus results from the inferior quality of inputs (or lack of skills in applying them) rather than a lack of incentive to allocate inputs to mixed crop-livestock farming. For this reason we find no empirical basis to support the hypothesis that land tenure is a constraint to agricultural productivity. © 1999 Elsevier Science B.V. All rights reserved.