<p><strong>Abstract.</strong> Nares Strait in northwest Greenland is one of the main gateways for oceanic freshwater and heat exchanges between the Arctic and the North Atlantic. With a changing Arctic climate, understanding the processes that govern the oceanic circulation in Arctic straits has become crucial and urgent, but this cannot be done with current geographically and temporally sparse in-situ observations only. High resolution regional modelling is thus required, but costly. We here report on one-year sensitivity experiments performed with the coupled ice-ocean regional model MITgcm to determine the relative importance of wind forcing, initial stratification and sea ice thickness on the accuracy of the modelled oceanic circulation in Nares Strait. We find that the modelled basin's circulation is mainly driven by density gradients in the upper oceanic layer, making accurate initial fields of temperature and salinity essential for a realistic oceanic circulation. The influence of the wind and sea ice thickness is less important, potentially making such high resolution fields not necessary for accurate strait modelling, provided these results are valid for other sea ice models as well. Comparison with ship-based measurements collected in summer 2015 reveals the experiments to be too cold at the surface, probably because of a not-dynamic-enough sea ice cover. Although the modelled freshwater is rather accurate, large efforts need to be put into observing the ocean and the sources of freshwater continuously throughout the year to produce realistic and efficient model simulations of the Arctic Straits, key players in the entire Arctic system and global climate.</p>