The Agricultural Model Intercomparison and Improvement Project (AgMIP) is an international initiative that links climate, crop, and economic modeling communities with advanced information technology to improve crop and economic models and produce next-generation climate impact projections for agriculture. AgMIP aims to significantly enhance the characterization of world food security under climate change and improve adaptation capacity in both developing and developed countries. The project involves coordinated research activities across climate, crop modeling, economics, and information technology teams, along with cross-cutting themes addressing uncertainty, aggregation and scaling, and the development of Representative Agricultural Pathways (RAPs) to test climate change adaptations in the context of other regional and global trends. AgMIP's scientific approach includes intercomparisons of various modeling methods to improve crop and economic models and ensemble projections. The project focuses on understanding the impacts of climate variability and change on agriculture, requiring a transdisciplinary effort to consistently link state-of-the-art climate scenarios to crop and economic models. Crop model outputs are aggregated as inputs to regional and global economic models to determine regional vulnerabilities, changes in comparative advantage, price effects, and potential adaptation strategies in the agricultural sector. Pilot results demonstrate AgMIP's role in assessing climate impacts with explicit representation of uncertainties in climate scenarios and simulations using crop and economic models. An intercomparison of wheat model simulations near Obregón, Mexico reveals inter-model differences in yield sensitivity to CO₂ with model uncertainty holding approximately steady as concentrations rise, while uncertainty related to choice of crop model increases with rising temperatures. Wheat model simulations with mid-century climate scenarios project a slight decline in absolute yields that is more sensitive to selection of crop model than to global climate model, emissions scenario, or climate scenario downscaling method. A comparison of regional and national-scale economic simulations finds a large sensitivity of projected yield changes to the simulations' resolved scales. Finally, a global economic model intercomparison example demonstrates that improvements in the understanding of agriculture futures arise from integration of the range of uncertainty in crop, climate, and economic modeling results in multi-model assessments. AgMIP's structure includes four project teams (Climate, Crop Modeling, Economics, and Information Technology), with guidance provided by a Leadership Team, Steering Group, and Donor Forum. The project also includes three cross-cutting themes: Uncertainty, Aggregation and Scaling, and Representative Agricultural Pathways (RAPs). These themes address key challenges in model intercomparisons, uncertainty assessment, and the development of RAPs to enable testing of climate change adaptations in the context of other regional and global trends. AgMIP protocols provide guidance on the operating procedures, progress evaluations, and anticipated deliverables from each team and for the integration of the project as a whole. The protocols also describe the processes and tasks necessary to conduct internally consistent model intercomparisons and multi-model assessments efficiently and comprehensively.The Agricultural Model Intercomparison and Improvement Project (AgMIP) is an international initiative that links climate, crop, and economic modeling communities with advanced information technology to improve crop and economic models and produce next-generation climate impact projections for agriculture. AgMIP aims to significantly enhance the characterization of world food security under climate change and improve adaptation capacity in both developing and developed countries. The project involves coordinated research activities across climate, crop modeling, economics, and information technology teams, along with cross-cutting themes addressing uncertainty, aggregation and scaling, and the development of Representative Agricultural Pathways (RAPs) to test climate change adaptations in the context of other regional and global trends. AgMIP's scientific approach includes intercomparisons of various modeling methods to improve crop and economic models and ensemble projections. The project focuses on understanding the impacts of climate variability and change on agriculture, requiring a transdisciplinary effort to consistently link state-of-the-art climate scenarios to crop and economic models. Crop model outputs are aggregated as inputs to regional and global economic models to determine regional vulnerabilities, changes in comparative advantage, price effects, and potential adaptation strategies in the agricultural sector. Pilot results demonstrate AgMIP's role in assessing climate impacts with explicit representation of uncertainties in climate scenarios and simulations using crop and economic models. An intercomparison of wheat model simulations near Obregón, Mexico reveals inter-model differences in yield sensitivity to CO₂ with model uncertainty holding approximately steady as concentrations rise, while uncertainty related to choice of crop model increases with rising temperatures. Wheat model simulations with mid-century climate scenarios project a slight decline in absolute yields that is more sensitive to selection of crop model than to global climate model, emissions scenario, or climate scenario downscaling method. A comparison of regional and national-scale economic simulations finds a large sensitivity of projected yield changes to the simulations' resolved scales. Finally, a global economic model intercomparison example demonstrates that improvements in the understanding of agriculture futures arise from integration of the range of uncertainty in crop, climate, and economic modeling results in multi-model assessments. AgMIP's structure includes four project teams (Climate, Crop Modeling, Economics, and Information Technology), with guidance provided by a Leadership Team, Steering Group, and Donor Forum. The project also includes three cross-cutting themes: Uncertainty, Aggregation and Scaling, and Representative Agricultural Pathways (RAPs). These themes address key challenges in model intercomparisons, uncertainty assessment, and the development of RAPs to enable testing of climate change adaptations in the context of other regional and global trends. AgMIP protocols provide guidance on the operating procedures, progress evaluations, and anticipated deliverables from each team and for the integration of the project as a whole. The protocols also describe the processes and tasks necessary to conduct internally consistent model intercomparisons and multi-model assessments efficiently and comprehensively.