In Brazil, as much as 50% of the native tropical savannas and woodlands (known as the Cerrado) have been converted to farmland under intensive agriculture. Land cover change at this scale has important implications for water and energy budgets, but at present, these implications are poorly understood. A promising new technology for elucidating these relationships at large scales is remote sensing; but results obtained by this method need to be verified with groundâbased measurements. For this study, we used remote sensing to estimate changes in energy balance and evapotranspiration (ET) in SĂŁo Paulo State following conversion of native Cerrado woodlands to cultivated land for sugar cane and eucalyptus. We applied the surface energy balance algorithm for land (SEBAL) to ten TMâLandsat 5 images taken during 2005 at two locations (sugar cane fields and Cerrado woodlands). The SEBAL ET results were compared with ET data obtained via eddy covariance. There was good agreement between the two (R2â=â0·87), with no systematic overestimation or underestimation by SEBAL. Our SEBAL analysis indicated that land cover change causes strong shifts in albedo, landâsurface temperature, and vegetation index â all of which affect energy balance and ET. Both the SEBAL and eddy covariance measurements indicate that ET for sugar cane is only about 70% of that for native woodlands. SEBAL estimates found that, for cultivated eucalyptus plantations, ET was about 88% of that for Cerrado forests. Our results indicate that the SEBAL methodology is a reliable tool for estimating energy budgets and ET in Cerrado landscapes and, further, that largeâscale conversion of Cerrado woodlands to agricultural land significantly alters energy budgets and ET. Copyright © 2014 John Wiley & Sons, Ltd.