In this article, we quantify the impact of the crop diversification measure implemented in France as part of the 2013 Common Agricultural Policy (CAP) greening reform. We exploit a discontinuity in the constraints imposed on farms larger and smaller than 30 ha, respectively, and apply regression differences-indifferences with a regression discontinuity setup on land use data collected from a representative sample of French farmers. We find that farms greater than 30 ha increased compliance with the measure and the number of crops grown on their lands and that farms larger and smaller than 30 ha responded differently to the reform.