For the last decade, the search for suitable alternatives for Li-ion batteries (LIBs) has steeply increased. While elements such as lithium and cobalt, are in high demand for LIBs, they could soon be replaced by less costly and more abundant elements1. One attractive alternative is potassium-ion batteries (KIBs) because of the abundance of potassium resources, and the lower standard electrode potential of K/K+ compared to lithium2.For characterization of KIB materials can be used same approaches as for LIB. An exception to that is describing electrochemical behavior, especially in a half-cell. One of the most important challenges for KIBs is a reliable 3-el setup with a stable and reproducible reference electrode (RE). Without a RE, the only possibility to independently quantify the anode and/or cathode impedance is the disassembly of multiple cells and the recombination of the identical anode or cathode pairs in symmetrical cells3. While using a metallic reference electrode is a practical solution for LIBs (using Li metal), the high reactivity of K renders the metal unsuitable as a reference electrode in KIBs.In this study, we outline our research on stable and reliable reference electrodes and compare several established approaches, specifically K-metal, a partially discharged Prussian white (K2Fe[Fe(CN)6]) and a more classic Ag/AgCl redox couple. Except for battery research, Ag|AgCIsat REs are commonly used in non-aqueous media in electrochemistry and offer the advantage of comparatively high chemical inertness, unlike K-metal or cathode-type RE. It was found that the evolution of degradation products from K-metal also induced significant drifts on a partly charged cathode RE (Fig. a), thus excluding both as reliable RE systems.In our revised cell design, the Ag|AgCIsat reference electrode was integrated into a thin-layer battery compartment. With this 3-electrode configuration half and full cell experiments, previous 2-electrode tests were revisited. So, the only way to compare the behavior of electrodes in half in full is to have stable and chemically passive RE which is Ag/AgCl. With this setup we were able to study the impact of electrolyte additives on the electrode reactions in more detail and pinpoint cut-off limits in graphite/ K2Fe[Fe(CN)6]. For example, in our experiments it was observed how graphite electrodes degrade under the influence of the DTD (1,3,2-Dioxathiolane 2,2-Dioxide) additive(Fig b).