Previous empirical literature suggests that agricultural subsidies are capitalized into farmland rents and that the introduction of the 2003 decoupling reform of the EU Common Agricultural Policy, attaching the subsidy to land only, might have even extended the phenomenon of capitalization. Employing the FADN dataset for Italy we investigate this issue using methodologies accounting for selectivity, endogeneity and individual heterogeneity simultaneously. The evidence suggests that selectivity bias causes inconsistent estimation of parameters and wrong inference. Results reveal instead that, in Italy, there is no incidence of both coupled and decoupled payments.