Greater diversity (i.e., variety, balance and disparity) within agricultural systems is often suggested as a solution to promote redundancy within such systems and therefore increase their adaptive capacity and reduce their vulnerability against climate change and variability. Yet this assumption relies upon the gathering and integration of field- and herd-scale results at the farm scale. We have conducted a farm-scale simulation-based study to evaluate the potential for increasing adaptive capacity and reducing vulnerability of livestock systems to weather variability through increase of their agricultural diversity. We manipulate in the simulation, factors of change in the livestock systems resulting in less diverse and more diverse livestock systems to be simulated. We assume, a priori, that certain combinations of the factored system components bring redundancy in the system and in turn increase its adaptive capacity and reduce its vulnerability against weather variability. Simulated factors of change are: F1: a change in the ratio of the area mechanically harvested (vs. grazed) to the whole farm area; F2: a change in the crops and grassland types grown or in the distribution of the area between crops and/or grassland types; F3: a change in calving periods from one season to another. The simulation plan includes a baseline scenario without changes and scenarios corresponding to all possible combinations of F1-F3. These scenarios are applied to four livestock systems located on a diagonal across France over a succession of four years with varying weather conditions. In these systems, self-sufficiency for forage is jeopardized by unfavorable years, and this may increase animal feeding costs. Thus we consider that adaptive capacity increases and vulnerability decreases as long as self-sufficiency for forage is achieved without increasing animal feeding costs. Results confirm the potential for increasing adaptive capacity and reducing vulnerability of livestock systems to weather variability through increase of their agricultural diversity. For instance, F2 has three main kinds of impacts on self-sufficiency for forage: (i) it yields significant average improvements by 34%, 43%, 36% and 36% across livestock systems for the four successive years, (ii) it buffers year-to-year variations and (iii) the final level of self-sufficiency is higher than the initial one even after two years with unfavorable weather conditions. Moreover, simulated changes do not increase animal feeding costs. Thus our results provide empirical evidence at the farm scale to supplement literature reviews based on field- and herd-scale results. They also confirm that through easily implementable on-farm changes, adaptive capacity can be increased and vulnerability of agricultural systems to weather variability decreased.