Cathodes based on layered LiMO2 are the limiting components in the path toward Li-ion batteries with energy densities suitable for electric vehicles. Introducing an overstoichiometry of Li increases storage capacity beyond a conventional mechanism of formal transition metal redox. However, the role and fate of the oxide ligands in such intriguing additional capacity remain unclear. This reactivity was predicted in Li3RuO4, making it a valuable model system. A comprehensive analysis of the redox activity of both Ru and O under different electrochemical conditions was carried out, and the effect of Li/Ru ordering was evaluated. Li3RuO4 displays highly reversible Li intercalation to Li4RuO4 below 2.5 V vs Li+ /Li0 , with conventional reactivity through the formal Ru5+−Ru4+ couple. In turn, it can also undergo anodic Li extraction at 3.9 V, which involves O states to a much greater extent than Ru. This reaction competes with side processes such as electrolyte decomposition and, to a much lesser extent, oxygen loss. Although the associated capacity is reversible, reintercalation unlocks a different, conventional pathway also involving the formal Ru5+−Ru4+ couple despite operating above 2.5 V, leading to chemical hysteresis. This new pathway is both chemically and electrochemically reversible in subsequent cycles. This work exemplifies both the challenge of stabilizing highly depleted O states, even with 4d metals, and the ability of solids to access the same redox couple at two very different potential windows depending on the underlying structural changes. It highlights the importance of properly defining the covalency of oxides when defining charge compensation in view of the design of materials with high capacity for Li storage.