The discovery of Lytic Polysaccharide Monooxygenases (LPMOs) has been instrumental for the development of economically sustainable lignocellulose biorefineries. Despite the obvious importance of these exceptionally powerful redox enzymes, their mode of action remains enigmatic and their activity and stability under process conditions are hard to control. By using enzyme assays, mass spectrometry and experiments with labeled oxygen atoms, we show that H2O2, and not O2 as previously thought, is the co-substrate of LPMOs. By controlling H2O2 supply, stable reaction kinetics and high enzymatic rates are achieved, the LPMOs work under anaerobic conditions, and the need for adding stoichiometric amounts of reductants is alleviated. These results offer completely new perspectives regarding the mode of action of these unique mono-copper enzymes, the enzymatic conversion of biomass in Nature, and industrial biorefining.