Ruthenium PNP complex 1a (RuH(CO)Cl(HN(C2H4Pi-Pr2)2)) represents a state-of-the-art catalyst for low-temperature (<100 °C) aqueous methanol dehydrogenation to H2 and CO2. Herein, we describe an investigation that combines experiment, spectroscopy, and theory to provide a mechanistic rationale for this process. During catalysis, the presence of two anionic resting states was revealed, Ru–dihydride (3–) and Ru–monohydride (4–) that are deprotonated at nitrogen in the pincer ligand backbone. DFT calculations showed that O- and CH- coordination modes of methoxide to ruthenium compete, and form complexes 4– and 3–, respectively. Not only does the reaction rate increase with increasing KOH, but the ratio of 3–/4– increases, demonstrating that the “inner-sphere” C—H cleavage, via C—H coordination of methoxide to Ru, is promoted by base. Protonation of 3– liberates H2 gas and formaldehyde, the latter of which is rapidly consumed by KOH to give the corresponding gem-diolate and provides the overall driving force for the reaction. Full MeOH reforming is achieved through the corresponding steps that start from the gem-diolate and formate. Theoretical studies into the mechanism of the catalyst Me-1a (N-methylated 1a) revealed that C—H coordination to Ru sets-up C—H cleavage and hydride delivery; a process that is also promoted by base, as observed experimentally. However, in this case, Ru–dihydride Me-3 is much more stable to protonation and can even be observed under neutral conditions. The greater stability of Me-3 rationalizes the lower rates of Me-1a compared to 1a, and also explains why the reaction rate then drops with increasing KOH concentration.