We found a systematic mismatch of modeled and measured NH₃ fluxes using two state-of-the-art dry deposition models and data from five sites in Europe. Results of our analysis indicate a too large minimum non-stomatal resistance and too strong temperature response in the parameterization of a unidirectional surface–atmosphere exchange scheme, as well as room for improvement on the emission potential parameterization of a bidirectional model, both directly impacting predicted NH₃ exchange fluxes.