A novel approach of modeling gas-particle partitioning using polyparameter linear free energy relationships (ppLFERs) is implemented into three different multimedia box models (the OECD Pov and LRTP screening tool (The Tool), ChemRange, and CliMoChem) and compared to approaches based on the octanol-air partition coefficient (K(OA)). In all three multimedia models, calculated overall persistence is not strongly influenced (differences < 3%) by the gas-particle partitioning approach selected. The long-range transport potential (LRTP) is more sensitive to the choice of the gas-particle partitioning model. In CliMoChem, the LRTP of polar chemicals is higher if the ppLFER gas-particle partitioning approach is applied, with differences up to a factor of 2. Modeled concentrations of polar chemicals in Arctic air are also higher in the ppLFER version of CliMoChem. The model results obtained with the ppLFER approach are in good agreement with measured concentrations of alpha-HCH, methoxychlor, and trifluralin in Arctic air, whereas results from the K(OA)-based version of the model are in some cases lower by a factor of 10-100. If the required chemical property data are available, the ppLFER approach holds considerable potential to improve the gas-particle partitioning description for polar chemicals in multimedia models.