The first phase of the PlioMIP shows that, while many characteristics of reconstructed sea surface temperature (SST) of the mid-Pliocene warm period (MPWP) are reproduced by climate models, there is a persistent underestimation of the reconstructed North Atlantic Ocean SST polewards of 30°N in the models (Dowsett et al., 2013), hampering our understanding of climatic mechanisms that allowed relative warmth of the mid- to high-latitude North Atlantic Ocean during the MPWP. While many explanations for a model-reconstruction-mismatch are possible, for example a mixed- seasonality signal in the MPWP reconstruction (Dowsett et al., 2013), we explore here the effect of the applied tectonic boundary condition. We consider uncertainty with respect to the configuration of the Bering Strait (open/closed) during the MPWP. The Bering Strait opened at about 5.3 Million years ago (Myr BP, Gladenkov et al., 2002) and saw a through-flow reversal around 3.6 Myr BP, after which low-saline North Pacific Ocean water may have decreased Atlantic Ocean overturning circulation (Matthiessen et al., 2009), potentially causing a cooling of the North Atlantic Ocean. Based on a comparison of the root mean square deviation between reconstructed MPWP SST and 1) a MPWP climate simulation with open Bering Strait (as generally assumed in the PlioMIP) and 2) a MPWP climate simulation with closed Bering Strait, we show that a closed Bering Strait leads to better agreement of a Community Climate System Models (COSMOS) simulation with the reconstruction, particularly in North Atlantic Ocean regions that are very warm during the MPWP. We propose to consider sensitivity studies in PlioMIP phase 2 that shall test whether simulations with other models lead to a similar inference. We plan to test whether a closed Bering Strait may remove a part of the model-reconstruction mismatch also in other models. To this end we will employ two additional climate model setups, encompassing the successor of COSMOS (ECHAM6- MPIOM; Jungclaus et al., 2013) and ECHAM6 coupled to an alternative finite element ocean model (ECHAM6-FESOM, Sidorenko et al., 2014), the latter being able to better resolve the critical coastal regions at northern high latitudes. These experiments provide a deeper understanding of the effect of gateway changes on MPWP climate, and allow us to explore the phase space of possible solutions of MPWP climate by using different model versions. As a logical next step we will examine the feedbacks employing an interactive ice sheet in our climate model, and plan to perform sensitivity studies that quantify the climatic impact of orbital variations for different settings of an open and closed Bering Strait.