This study investigates the origin of electrocatalytic oxygen reduction activity in graphene-based catalysts, aiming to achieve optimal performance. The research combines experimental and computational approaches to understand the factors influencing the catalytic activity of doped graphene. Experimental data includes characterization of catalysts, electrochemical measurements, and analysis of ORR activity. Computational methods involve model building, reaction pathway analysis, and free energy calculations to determine the most favorable sites for oxygen reduction. The study identifies that the associative 4e⁻ reduction pathway is more favorable than the dissociative pathway for oxygen reduction on doped graphene due to the high energy barriers for O₂ dissociation. The most active sites for ORR are determined based on the adsorption free energy of intermediates such as OOH*, O*, and OH*. The study evaluates various doped graphene structures, including boron-doped, nitrogen-doped, oxygen-doped, phosphorus-doped, and sulfur-doped graphenes, to identify the most effective catalysts. Key findings indicate that the pyridinic nitrogen-doped graphene (pdN-G) and pyO-G (pyran type oxygen-doped graphene) exhibit the lowest reaction barriers and highest ORR activity. The active sites for these materials are identified as specific carbon atoms adjacent to the heteroatoms. The study also highlights the importance of the adsorption of intermediates and the activation energy of reaction pathways in determining the catalytic efficiency of graphene-based materials. The results provide a roadmap for designing high-performance graphene-based catalysts for oxygen reduction reactions.This study investigates the origin of electrocatalytic oxygen reduction activity in graphene-based catalysts, aiming to achieve optimal performance. The research combines experimental and computational approaches to understand the factors influencing the catalytic activity of doped graphene. Experimental data includes characterization of catalysts, electrochemical measurements, and analysis of ORR activity. Computational methods involve model building, reaction pathway analysis, and free energy calculations to determine the most favorable sites for oxygen reduction. The study identifies that the associative 4e⁻ reduction pathway is more favorable than the dissociative pathway for oxygen reduction on doped graphene due to the high energy barriers for O₂ dissociation. The most active sites for ORR are determined based on the adsorption free energy of intermediates such as OOH*, O*, and OH*. The study evaluates various doped graphene structures, including boron-doped, nitrogen-doped, oxygen-doped, phosphorus-doped, and sulfur-doped graphenes, to identify the most effective catalysts. Key findings indicate that the pyridinic nitrogen-doped graphene (pdN-G) and pyO-G (pyran type oxygen-doped graphene) exhibit the lowest reaction barriers and highest ORR activity. The active sites for these materials are identified as specific carbon atoms adjacent to the heteroatoms. The study also highlights the importance of the adsorption of intermediates and the activation energy of reaction pathways in determining the catalytic efficiency of graphene-based materials. The results provide a roadmap for designing high-performance graphene-based catalysts for oxygen reduction reactions.