1 Introduction Over the past 50 years, agrifood systems in France and in Europe have been evolving towards a simplification and a regional specialization of crops and cropping systems (Schott et al., 2010). This evolution was originally driven by goals of mass production of food and economies of scale, but its sustainability is now questioned. However, reversing the trend from simplification towards diversification requires redesigning agrifood systems as a whole, considering each component of these systems and their interactions with the other components (Hill and MacRae, 1996; Meynard et al., 2018). While the lock-in effects affecting current agrifood systems have already been studied, there is a lack of literature on the levers that can be mobilised by actors to change their practices towards crop diversification. Some studies (Brodt et al., 2004; Buck et al., 2001; Greiner et al., 2009) provide answers as to why some farmers decide to change their practices towards more sustainable or agroecological practices, but they do not examine how they are able to implement these changes, or how they interact with their sociotechnical environment as they change (Blesh and Wolf, 2014). Compared to other changes in practices such as input reduction or conversion to no-till systems, crop diversification is not a goal concerning only farmers, as it is strongly dependent on the practices of actors outside the farming system. We hypothesize that analysing how farmers change their practices towards crop diversification, in relation with other actors from the agrifood system, will enable us to better understand the drivers of farmers choices when they diversify their crops (why?), and to identify levers for implementing crop diversification more largely in farming and agrifood systems (how?). 2 Materials and Methods We carried out in-depth surveys with farmers that have changed their practices towards crop diversification, in a French region where several minor crops coexist among a few major crops. We traced with these farmers the diversification process they had carried out over time, and we questioned over this process (i) how their cropping systems and the technical management of their crops evolved, (ii) how they took their decisions before, during and after implementing the changes, (iii) what productive resources (land, labour, machinery, inputs), knowledge and learning processes they mobilized, and (iv) how they marketed the new crops. We paid special attention to the influence of the interactions of farmers with other actors from their sociotechnical environment (other farmers, value-chain actors, extension services) on this diversification process. 3 Results This research provides contrasted examples of farmers’ processes of change towards more diversified cropping systems, including a wide range of protein, fibre, grain and oilseed crops. Our first results show that crop diversification at the farm level is triggered by technical or economic problems encountered on at least one of the two main crops in local cropping systems, wheat and maize. Farmers who diversify their crops generally replace some or all of their maize area by new crops, but they do not call into question the position of wheat in their cropping systems. For more than half of the farms in our sample, the crop diversification process follows a similar, “cautious” pattern: during the first steps of the process, lasting 6 to 9 years, the farmers progressively increase the number of “diversification” crops (i.e. crops other than wheat and maize) and their share in the utilised agricultural area (UAA), until this share reaches a threshold around 50% of the UAA, with 3 to 6 diversification species. The diversification species and their respective share are then adjusted by the farmers within this 50% threshold. The choice of the new species to be cultivated is strongly constrained by soil characteristics and by outlet opportunities specific to each farm. During the diversification process, most farmers develop strategies in order not to modify in-depth their own workforce, equipment or land: they rather mobilize external resources such as contractors or common machinery. Moreover, crop diversification is often considered by farmers as a way to reduce their workload, especially with crops that present a growth cycle different from main crops and that require few interventions. Finally, we observed that farmers’ performance criteria for crop diversification evolve during the diversification process, as they learn about new crops and cropping systems, to include more and more pre-crop and multi-year effects. For example, farmers will observe that a crop such as hemp improves the soil structure and decreases weed populations for the subsequent crop. These unforeseen pre-crop benefits can be integrated in farmers’ assessment of the new crops. As a consequence, these crops will be maintained in the system even if their economic performances – assessed through annual gross or semi-net margins – appear lower than for other crops. 4 Discussion and Conclusions This research highlights that crop diversification, at the farm level, is a complex and long-term process which requires farmers to re-design not only their cropping systems, but also their indicators for assessing these cropping systems, and how they mobilise productive resources. Though each farm follows a specific crop diversification process, leading to a specific cropping system, we were able to identify some common patterns of crop diversification processes among farms. A complementary analysis focused on the evolutions of practices of other stakeholders of the region, involved with the farmers in the diversification process, should provide us with a global view of contrasted strategies of crop diversification, and their respective strengths and drawbacks. Our methods will also be replicated to the farms of different regions and contexts in Europe in order to compare our results with varied situations and to identify more generic levers for crop diversification. References Blesh, J., Wolf, S.A., 2014. Transitions to agroecological farming systems in the Mississippi River Basin: toward an integrated socioecological analysis. Agric Hum Values 31, 621–635. https://doi.org/10.1007/s10460-014-9517-3 Brodt, S., Klonsky, K., Tourte, L., Duncan, R., Hendricks, L., Ohmart, C., Verdegaal, P., 2004. Influence of farm management style on adoption of biologically integrated farming practices in California. Renewable Agriculture and Food Systems 19, 237–247. https://doi.org/10.1079/RAFS200488 Buck, A.J. de, Rijn, I. van, Roling, N.G., Wossink, G.A.A., 2001. Farmers’ reasons for changing or not changing to more sustainable practices: An exploratory study of arable farming in the Netherlands. The Journal of Agricultural Education and Extension 7, 153–166. https://doi.org/10.1080/13892240108438817 Greiner, R., Patterson, L., Miller, O., 2009. Motivations, risk perceptions and adoption of conservation practices by farmers. Agricultural Systems 99, 86–104. https://doi.org/10.1016/j.agsy.2008.10.003 Hill, S.B., MacRae, R.J., 1996. Conceptual Framework for the Transition from Conventional to Sustainable Agriculture. Journal of Sustainable Agriculture 7, 81–87. https://doi.org/10.1300/J064v07n01_07 Meynard, J.-M., Charrier, F., Fares, M., Le Bail, M., Magrini, M.-B., Charlier, A., Messéan, A., 2018. Socio-technical lock-in hinders crop diversification in France. Agron. Sustain. Dev. 38, 54. https://doi.org/10.1007/s13593-018-0535-1 Schott, C., Mignolet, C., Meynard, J.-M., 2010. Les oléoprotéagineux dans les systèmes de culture : évolution des assolements et des successions culturales depuis les années 1970 dans le bassin de la Seine. OCL 17, 276–291. https://doi.org/10.1051/ocl.2010.0334