Farms in the US Midwest are experiencing disruptive heavy rainfall during the spring planting window and more intense droughts during the summer growing season. Recent analyses of field experiments suggest that increasing crop rotational complexity (RC) builds resilience to climate change, especially when rotations include overwintering cover crops that assimilate and retain nutrients and increase organic matter inputs to soil. However, evidence from working farms is lacking. We used field-scale remote sensing to assess relationships between crop diversification and agroecosystem climate resilience from 2008 to 2019 in Michigan, the most agriculturally diverse state in the US Midwest. Using panel fixed effects models and linear regressions, we assessed how corn and soybean yield and temporal yield stability (measured as the coefficient of variation) respond to: (1) an index of crop RC that encompasses species turnover and functional diversity, and (2) replacing bare fallows with winter cover crops. We also tested how cover crops influence corn and soybean planting dates as an indicator of resilience to heavy spring rainfall. We found that crop diversification was linked to higher corn and soybean yields, but cover crops showed more promise for improving yield stability than overall RC. Cover crops reduced planting delays associated with heavy spring rainfall, and did not interfere with primary crop planting under average weather conditions. However, cover crop benefits for resilience took multiple years to appear on farms, highlighting the importance of technical and financial support during the early phases of transition. Together these findings may promote cover crop adoption by reducing concerns that cover crops interfere with primary crop planting, and instead showing that they are a crucial tool for minimizing risk from increasingly variable and extreme climatic conditions.