We study the effect of the Agenda 2000 CAP reform on the choice of crop mix with a differences-in-differences regression model. In our analysis, we exploit a quasi-experimental setting between Finnish agricultural subsidy regions that emerged in the reform. The variation catches the causal effect between coupled subsidies and planting decision. We found that changes in payments to individual crops explain changes in crop mixes. However, the observed causal effect does not fully conform to theoretical predictions. Additionally, the effect is small and is easily confounded by small changes in natural conditions. The results indicate that changes even in coupled subsidies may only have small impact on production decisions as yet unexplored factors prevent the incidence from being perfect. Acknowledgement : I want to thank my supervisors Dr. Elias Eini and Reaserch Director Marita Laukkanen for many useful advices on this paper.