As part of the EU-ECLIPSE project, a set of global and regional chemistry and transport models have performed simulations predicting atmospheric distributions of aerosols both globally and in the Arctic region. Here, model accuracy has been evaluated using satellite observations in the Arctic. All models used the same anthropogenic (ECLIPSEv4a), biogenic (MEGAN), and biomass burning (GFED-v3) emissions. Aerosol data derived from the spaceborne CALIOP lidar are used to evaluate models over the entire Arctic region during spring and summer 2008. Monthly mean results from the ECLIPSE models are compared with vertical distributions of attenuated scattering ratio at 532 nm from CALIOP averaged using a 3°x5° grid over a region covering 30° N to 82° N. This new product (Ancellet et al., 2014), based on CALIOP level 1 backscatter signals filtered using CALIPSO level 2 cloud masks, is not affected by errors associated with a lidar signal inversion. Modelled aerosol distributions are shown to overestimate aerosols in mid-latitude source regions and underestimate aerosols in the lower Arctic troposphere (0-2km), compared to CALIOP satellite data despite the fact that models use the new ECLIPSE emissions that include improved representations of Russian flaring and domestic wood burning. . During April 2008, lower aerosol levels are observed over the Arctic near the surface compared to the free troposphere. Intense fires over Siberia and western Russia/eastern Europe coupled with transport of Asian emissions into the Arctic free troposphere during spring 2008 led to enhanced aerosol concentrations aloft at this time of year. The models have problems simulating enhanced aerosols aloft during spring 2008. In general, observed aerosol levels are lower during the summer months due to more efficient scavenging processes and less efficient transport of polluted air masses into the Arctic at this time of year. All the ECLIPSE models have too little aerosol in the Arctic in July 2008 pointing to aerosol lifetimes that are too short or suggesting that vertical transport in the models may not be efficient enough. Discrepancies point to remaining uncertainties in emissions as well as deficiencies in global model transport processes and uncertainties in wet scavenging of aerosols during transport to the Arctic.