Due to climate change, farms must be able to withstand increased weather volatility. Crop diversification as a strategy to enhance resilience in the face of extreme weather events is explored in the context of northeastern dairy cropping systems. To address resilience, the farm is understood using a social-ecological systems perspective. In the industrialized setting, the farm as a land use system overlaying other biotic communities is embedded in a socio-technical configuration consisting of individual and institution actors who enforce/reinforce norms, rules, practices, and technologies. Operationalized, resilience is the interaction of ecological, economic, and social elements that produce and restore ecosystem services. Focusing on a farmer’s management, the cross-scale feedback mechanisms in the complex system are the criteria of innovation diffusion: relative advantage, compatibility, triability, observability, and complexity. Crop diversification provides farmers with a relative advantage as it increases the adaptability of the farmer through a diversity of practices, and at the field-level, increases the spatial and temporal variance which improves the ability to absorb stress. Thus, the benefits of diversification are derived from increased complexity, which often inhibits diffusion. Compounding this, the socio-technical configuration in which the farmer is embedded shapes their behavior and masks ecological signals as seen through 25 semi-structured interviews with conventional and organic dairy farmers. Increasing adoption of crop diversification practices requires creating agronomic knowledge that can be appropriately contextualized, increasing coherence in research findings for easier synthesis, and designing science for application to advance technologies and policies. I worked to this end in two research experiments. Crop diversification strategies were deployed in a multi-site, multi-year forage intercropping experiment, and productivity, weed suppression, forage quality, and yield stability were quantified. Intercropping practices employing interspecific diversity had a greater effect than practices employing intraspecific diversity on yield stability, likely due to an increase in functional and response diversity. I also constructed a framework to design crop mixtures using ecological models and a series of response-surface designs. The framework accounts for the different competitive traits of species so that crop mixtures are more able to deliver multiple ecosystem services. It can be used when recommending seeding rates for mixtures to farmers and may reduce some barriers of adoption.