We propose to analyze in detail the rapidly changing agricultural frontier in the state of Mato Grosso Brazil, a part of the Legal Amazon but encompassing both tropical forest and savannah natural landscapes. This region provides us a unique opportunity to study both one of the most rapidly changing agricultural frontiers in the world and two globally important biomes, using a phenological approach to characterizing land cover change from remotely sensed data over a longer period than was previously possible. The research consists of two primary themes: 1) detection and characterization of land cover change, and 2) explaining and attributing the observed changes. The proposed research will use MODIS data to detect and characterize land cover change, and moderate resolution sensors (e.g. Landsat) to assist the analyses and to aid in verification and validation. It builds on our previous successful application of these approaches to characterize land cover change over a shorter period. The new research will incorporate analyses of socioeconomic drivers of extensification (conversion of natural vegetation to human use) and intensification (conversion of pasture to row-crop or single- to double-cropping) at the county level using standard spatial econometric approaches and a dataset assembled during this research.