In order to build sustainable agricultural land systems, agronomist should adopt a landscape perspective by simulating 1) scenarios at the landscape scale, 2) farmers’ decision processes for cropping system choices, and 3) cropping systems performances at field scale. We here propose a modeling framework, based on a scenario approach with an optimization model to prototype sustainable land systems. The framework uses a regional bioeconomic model which integrates farmers’ decision process on the adoption of cropping systems at field scale. Farmers’ decisions are driven by a combination of spatially explicit biophysical conditions of their fields in a landscape and socio-economic condition of their farms in a regional and global context. The scenario method encompasses normative, exploratory, optimized scenarios to assess the potential of levers to improve the response of agricultural land systems to sustainability goals with scenarios. The framework was tested in Guadeloupe for defining a set of scenarios assessing several combinations of agronomic, technical, economic and social levers to build new agricultural land systems improving the contribution of agriculture to sustainable development. The pathway of scenario development proposed by the framework led to the definition of a multi-functional "Go sustainable" scenario in Guadeloupe that combined among others new cropping systems, with a reorientation of subsidies towards local productions, a more experienced workforce and a ban of food crops production on polluted soils. This scenario increased the food and energy self-sufficiency, the employment and decreased the risk of pollution in water bodies and the risk of contamination of crops due to polluted soils. This framework can be used to understand the potential of response to sustainability issues to help local decision makers in the definition and implementation of policies that take into account the diversity of farms and field conditions in the region.