It is well known that electrocatalytic oxygen evolution reaction (OER) activities primarily depend on the active centers of electrocatalysts. In some oxide electrocatalysts, high-valence metal sites (e.g., molybdenum oxide) are generally not the real active centers for electrocatalytic reactions, which is largely due to their undesired intermediate adsorption behaviors. As a proof-of-concept, molybdenum oxide catalysts are selected as a representative model, in which the intrinsic molybdenum sites are not the favorable active sites. Via phosphorus-modulated defective engineering, the inactive molybdenum sites can be regenerated as synergistic active centers for promoting OER. By virtue of comprehensive comparison , it is revealed that the OER performance of oxide catalysts is highly associated with the phosphorus sites and the molybdenum/oxygen defects. Specifically, the optimal catalyst delivers an overpotential of 287 mV to achieve the current density of 10 mA cm -2 , accompanied by only 2% performance decay for continuous operation up to 50 h. It is expected that this work sheds light on the enrichment of metal active sites via activating inert metal sites on oxide catalysts for boosting electrocatalytic properties.