How did technical change affect land use in Brazilian agriculture? We use data from the last two Agricultural Censuses of 1995/1996 and 2006 to answer this question for five different regions. We focus on the estimation of the Hicksian bias induced by technical change over this period and found that technical change was, in general, land-using. In the Southeast region, we found labor-saving behavior. Both results can be interpreted in light of the induced innovation hypothesis under Acemoglu s approach that allows testing when prices are not available. Acknowledgement :