Artificial photosynthesis could promise abundant, carbon-neutral energy, but implementation is currently limited by a lack of control over the multi-electron catalysis of water oxidation. Discoveries of the most active catalysts still rely heavily on serendipity. [Ru(tpy)(bpy)(H2O)](2+) (1; bpy = 2,2'-bipyridine, tpy = 2,2';6',2 ''-terpyridine) is representative of the largest known class of water oxidation catalysts. We undertook an extensive spectroscopic analysis of the prototypical 1 water oxidation catalyst and its fastest known analog [Ru(EtOtpy)(bpy)(H2O)](2+) (2), capable of 10 times faster water oxidation, to investigate the mechanism of action and factors controlling catalytic activity. EPR and resonance Raman spectroscopy did not detect the proposed [Ru-v=O] intermediate in 1 and 2 but indicated the possible formation of N-oxides. A lag phase was observed prior to O-2 evolution, suggesting catalyst modification before the onset of catalysis. The reactive intermediate [Ru(tpy)(bpy-NO)(H2O)](2+) (1-NO; bpy-NO = 2,2'-bipyridine-N-oxide) proposed by combined spectroscopic and DFT analysis was de novo synthesized and demonstrated 100-fold greater catalytic activity than 1. Thus, in situ transient formation of small amounts of the Ru complex with N-oxide ligands can significantly activate single-site Ru-based catalysts. Furthermore, the rate of O-2 evolution was found to correlate with the redox potential of the ligand. This observation might assist with rational design of new catalysts.