Climate variability is predicted to increase in sub-Saharan Africa. This will impact the food security of the populations who rely on rainfed cereal production. Sustainable intensification of cropping systems aims at increasing food production while adapting to future climate and mitigating environmental impacts. The opportunity to diversify cereal-based cropping systems with legumes to increase resilience has not been quantitively assessed so far. In this study, we explored the impact of interannual climate variability for a set of cropland allocations, going from maize only to various share and number of legume crops. We explored the hypotheses that maize and legumes inter-annual performances are not correlated, so that mixing the two crops will increase the resilience of the cropland allocation. We calibrated the STICS soil-crop model for maize, cowpea, groundnut and pigeon pea, with data from on-farm trials in Murehwa district in sub-humid Zimbabwe, conducted in the 2021- 2022 and 2022-2023 growing seasons. Crop performance in terms of average yield and its variability was simulated with the calibrated crop model for the historical climate (1996-2016), on red-clay and granitic sandy soils, for more fertile homefields and less fertile outfields. Two planting dates (early-late) and three levels of fertilization (0-80-160 kg N/ha), brought by mineral fertiliser only, manure only, or by a combination of 30% manure and 70% mineral fertiliser, were investigated. The simulated crop performance was used to compute for each land allocation the agricultural productivity, the economic profitability and the environmental performance. From the on-going simulation results, we will draw critical insights on the complementarities between maize and legumes. More specifically, we expect smaller but more stable groundnut yield compared with fertilized maize, so that diversification with groundnut will increase production stability and food security in terms of proteins. With regar