Current and proposed policies aim to leverage the Federal Crop Insurance Program (FCIP) for adoption of conservation practices. This study use nationally representative field and farm-level data to inform the effectiveness of targeting FCIP participants. We implement an unsupervised machine learning model to assess what set of conservation practices are most common among farms that use crop insurance. Next, we introduce a novel approach to use survey data to measure nitrogen (N) balance, a yield-scaled measure of nitrogen fertilizer’s environmental impact. We then test then whether crop insurance predicts more optimal N Balance. We find that farms that use crop insurance may have higher adoption rates for conservation practices that are also generally profitmaximizing.