In agricultural landscapes, many environmental processes depend on spatial arrangement and dynamic of crops linked to farm management decisions. Given the complexity of managing both socio-technical and environmental processes, territorial design can use software simulation tools. Most of the crop allocation simulation tools are either not taking into account farm level, or take the farm level too precisely to be usable at the landscape level. We propose CAPFarm, a multi-year cover allocation simulator available at the farm level and also at the landscape level. CAPFarm allows users i) to formalize cover allocation decision rules into a system of spatial and temporal constraints and ii) to solve the system of constraints in order to find consistent solutions. First we present theoretical and applied aspects of CAPFarm. Next we show its functioning at both levels by using examples with environmental focuses. At the farm scale, we evaluate a scenario of practice changes by increasing the duration of temporary grasslands. At the landscape level, we try to increase the grasslands connectivity by exploring crop allocation diversity and by adding crop allocation constraint at the landscape level. Then, we show that CAPFarm is usable to detect issues of feasibility and is helpful to propose alternative scenarios. Finally, we discuss the relevance to use it in a territorial design way.