Vanadium redox flow battery (VRFB) is a promising technology for energy storage because of its independent energy to power ratio and long cycle life. However, VRFB commercialization is still hindered by some technological issues, among which the electrolyte imbalance caused by the undesired vanadium-ions and water transport through the membrane. These phenomena result in battery capacity loss, that can be recovered with periodic re-balancing procedures, thus increasing operating and maintenance costs.In the literature different studies have been dedicated to the analysis of ions and water crossover. In particular, Oh et al [1] evidenced that water and vanadium ions transport across the separator are strictly related. Shin et al [2] tried experimentally to mitigate water imbalance through a modified preparation of the electrolytes; however, the proposed mitigations were experimentally verified for just one charge-discharge cycle.In this work, combining experimental and modelling activity, the evolution of discharged capacity and electrolyte volume were first investigated and then mitigated modifying electrolytes composition for hundreds of cycles. In addition, to increase the validity of the obtained results, three different Solvay® membranes were tested: E98-05 (EW 980, thickness 50 mm), E87-05 (EW 870, thickness 50 mm) and E98-09 (EW 980, thickness 90 mm). Cycles with fixed cut-off voltages at 0.1 A cm-2 were performed in a 25 cm2 cell employing carbon felt electrodes (Sigracell® GFD 2.5 EA thermal activated).Adopting commercial electrolyte, the discharged capacity of the battery dramatically decreased in its first phase of operations, while electrolytes volume continuously varied until the end of the test, as reported in figure 1B for E98-05. This behaviour was showed by all the three different membranes, regardless their thickness or equivalent weight. Thanks to a preliminary modelling analysis with a 1D VRFB physical model, the causes of the initial capacity decay were fully understood and three different mitigation strategies obtained through the modification of electrolyte preparation were proposed and experimentally validated.The first one consisted in the reduction of the initial proton concentration difference between anolyte and catholyte, obtained by increasing the acid molarity at negative side. This led to the mitigation of the initial capacity decay. The second one consisted in the increase of the proton concentration in both the catholyte and anolyte, obtained by increasing their acid molarity. This led to the mitigation of the initial capacity decay and the reduction of the electrolyte volume variation in time.The third mitigation strategy combined the beneficial effect of reducing the initial proton concentration difference between anolyte and catholyte and the one of increasing the proton concentrations in both the electrolytes. The results were promising: as can be seen in figure 1, with E98-05 the initial capacity decay was completely mitigated and no volume variation was experienced in more than 200 cycles. The effectiveness of this mitigation strategy was experimentally validated successfully also with E98-09 and E87-05. Figure 1 - charge-discharge cycle with E98-05 using modified electrolytes: A) discharged capacity; B) electrolyte volume variation.