Agricultural expansion and intensification drive the conversion of natural areas worldwide. Scenarios are powerful tools to explore possible future changes in agricultural land use, how these may affect the environment, and how policies may influence land-use patterns. Focusing on Argentina’s prime agricultural areas, the Pampas, Espinal and Chaco, we developed spatially-explicit future land-use scenarios from 2010 to 2030, considering both agricultural expansion (i.e., conversions from woodland to either grazing land or cropland) and agricultural intensification (i.e., conversions from grazing land to cropland). Our simulations were based on an econometric model of net returns, which assumes economically rational land-use actors. Using this model, we assessed the rates and spatial patterns of future land-use change under current land zoning in our study region, and contrasted this with a forecast of future land use based on land-conversion rates from 2000–2010. We systematically tested the impact of economic policies (e.g., taxes or subsidies), infrastructure improvement (e.g., road paving), and technological innovation (i.e., yield increases) on the spatial patterns of land-use conversions. Our model suggests future land-use change will mainly happen along intensification pathways, with deforestation slowing down, if land-use actors would be profit-maximizing. This general pattern did not change even for policy interventions that impacted profits from agriculture in major ways, cautioning against overestimating the leverage that economic policies provide for halting deforestation. Improving the region’s road network would create a strong incentive to expand cropland further into remaining woodlands and over grazing lands. However, low agricultural profits and higher yields could curb deforestation in marginal areas to some extent. We also highlight that priority areas for conservation are particularly likely to experience high land-use pressure in the future. Given the lower-than-expected power of economic policies to alter deforestation patterns in our models, zoning, if properly enforced, appears to be a more straightforward tool for avoiding unwanted environmental impacts in the Chaco.