The electrochemical reduction of CO2 (eCO2R) is one promising attempt to create value added products from excess greenhouse gas with renewable energy. However, to become commercially relevant, some challenges have to be outgrown. To overcome the challenge of mass transport limitations on planar electrodes, Gas-Diffusion-Electrodes (GDE) have been implemented for eCO2R, allowing to reach sufficient reaction rates with the CO2 directly being delivered to the catalyst layer [1]. A major issue in the application of GDEs for eCO2R represents flooding of the mostly carbon based gas diffusion media by the alkaline electrolyte during application. This phenomenon results in a detrimental change of the 3-phase boundary and is accompanied by a loss of activity and a breakdown of performance [2]. A possible approach to avoid flooding of the GDE during eCO2R is to apply a backpressure in order to increase the capillary pressure inside the GDE, which needs to be balanced accurately, since excess backpressure can cause a CO2 crossover stream through the GDE into the electrolyte compartment [3]. In our work we investigate the permeation behavior of GDEs in a customized flow cell and investigate how to influence it via electrode preparation. A deeper understanding of the permeation characteristics of GDEs could lead to a better choice of operation points regarding flooding issues. However, the operation point itself and the preparation of GDEs could impact not only the flooding behavior of the electrode but also selectivity of the reaction. The eCO2R is taking place ~10 µm away from the gas-liquid boundary. The distribution of the two phases inside the GDE is crucial for the reaction pathway [4]. Therefore, we examine possible links of permeation behavior and electrode preparation on selectivity.In detail, our setup contains a customized flow cell in which the pressure in the gas compartment is adjusted with a needle valve and monitored via a pressure transmitter. The gas-crossover stream through the GDE is then registered by collecting the gas transferred to the liquid electrolyte in a measuring cylinder.For electrode preparation, commercial carbon based GDLs (25BC-Sigracet, H23C8-Freudenberg) are pretreated by either spray coating a carbon-ionomer-layer on top of the microporous layer or sputter coating with copper first. In the second step, a galvanostatic deposition of copper is carried out on the pretreated GDLs.The permeation behavior is tested for the non-coated commercial GDLs, the GDLs after the pretreatment step and after deposition of copper without and with current applied, respectively. In our experiments the crossover stream related to backpressure through the GDL H23C8 was lower compared to the crossover stream through GDL 25BC, most likely being linked to differences in the porosity and cracks of the MPL (Figure 1A). However, when comparing identically prepared GDEs based on different GDL types at the same crossover stream, only minor changes in selectivity can be observed (Figure 1B). It was also found that the crossover stream was influenced by the coating of the electrode. While sputter coating (spu) has a minor impact on gas permeation, spray coating (spr) leads to a significant change. These observations could be related to increased electrode wetting after spray- but not after sputter coating. The permeation behavior after galvanic deposition of copper is comparable, independent of the pretreatment procedure. Comparing the selectivity of GDEs with the different pretreatment procedure, a significant change in selectivity can be observed (Figure 1C). In the next steps, it will be crucial to investigate if this selectivity change is related to a different gas-liquid distribution in the GDE or to other factors as the composition of the coating layer to successfully influence operation stability and selectivity.