With the increasing scarcity of natural resources, the unsustainability of the conventional and intensive agriculture and the need of food security, agronomy engineering is facing a serious challenge. In this context, different projects aim to design and evaluate innovative cropping systems in order to increase the productivity of agro-ecosystems while preserving the diverse ecosystem services they provide or support. Simulation can help agronomists to test the robustness of field experiments results by testing various soil and climate conditions. While there are already crop models that simulate correctly crop development, there are few models that simulate the way farmer conduct their cropping systems. Yet, it is a key challenge because innovative cropping systems are often based on new farming practices, which take more into account the state of the crop and of the environment than conventional cropping systems. In addition, the use of “fixed dates” in simulations for farming practices is not adapted because it is important to take into account weather variations, which have a strong effect on the dates of farming operations. Therefore, the evaluation of innovative cropping systems by simulation requires building decision model that mimic farmers’ decision-making, and help in analyzing impacts of farmers’ practices on the sustainability of the cropping system. The modelling and simulation platform dedicated to the study of agro-ecosystems RECORD (J-E. Bergez et al., 2013) has been developed at INRA (French national institute for agricultural research). One of the objectives of the RECORD project was to help modelers in developing decision models and in coupling them to crop models. A generic conceptual decision-modelling framework (Bergez et al., 2016) has been proposed. It allows to design flexible management plan of activities using the concepts of a directed multigraph without loops and of a knowledge base. In the context of cropping system modelling, the graph of activities represents the farmer’s work plan and relies on the knowledge base to activate or disable technical operations. The knowledge base evolves all along the simulation collecting information provided by the biophysical model, as the farmer does when monitoring and observing the environment. Based on this conceptual decision-modelling framework an original graphical plugin “Decision” (Bergez et al., 2016) has been developed. It helps agronomist modelers in sketching and implementing their decision models and in linking them with biophysical models. The plugin allows defining activities (tasks), relation between activities and decision rules to trigger the different tasks. As the RECORD platform is based on DEVS (Discrete Event System Specification) formalism (Zeigler et al., 2000), the software implementation of the plugin is expressed in this formalism. We have applied this decision-modelling framework to the context of an ongoing project whose issue is bio economic evaluation of innovative cropping systems compared to conventional ones. To produce simulation results required by this project, a coupled model has been designed following the conceptual approach that an agricultural system can be divided into three sub-systems: Agent, Operating, and Biophysical (Le Gal et al. 2009; Martin-Clouaire and Rellier 2009). The model couples the crop model STICS (Brisson et al., 1998, J-E. Bergez et al., 2014), to a decision model using the Decision plugin and to a climate series reader. The model is generic because STICS can simulate the behavior of soil–crop systems for a large range of crops, and because the decision model is parametrable. It is used for different species commonly cultivated in south west of France (maize, sunflower, wheat, sorghum …), for different years (1982-2012), for two ways of farming practices: conventional (what do farmers commonly use) and innovative which are currently tested in experimental fields.