Integrated crop-livestock systems are a form of sustainable intensification of agriculture that rely on synergistic relationships between plant and animal components to bolster critical agroecosystem processes. We simulated cash crop and cover crop productivity dynamics in an integrated beef-soybean cover crop grazing system typical of southern Brazil to gain a better understanding of the impacts of livestock integration on system performance. Long-term historical simulations in APSIM showed that the integrated system resulted in greater system-wide productivity than a specialized control system in 77% of simulated years. Although soybean yields were typically lower in the integrated system, additional forage and livestock production was obtained. Under simulated future climate conditions (RCP8.5 scenario from 2020-2060), integrated system productivity exceeded specialized system productivity in 95% of years despite declines in average soybean yield and cover crop aboveground biomass production. While the integrated system provided a productivity buffer against chronic climate stress, its resilience to annual weather anomalies depended on disturbance type and timing. This study demonstrates the potential advantages of cover crop grazing as a sustainable intensification strategy, using process-based modeling tools to explore biophysical proxies for resilience in field-level agroecosystem functions.