We present the results of the first theoretical investigation of salen-manganese complexes as synthetic catalytic scavengers of hydrogen peroxide molecules that mimic catalase enzymes. Catalase mimics can be used as therapeutic agents against oxidative stress in treatment of many diseases, including Alzheimer's disease, stroke, heart disease, aging, and cancer. A ping-pong mechanism approach has been considered to describe the H2O2 dismutation reaction. The real compounds reacting with a peroxide molecule were utilized in our BP density functional calculations to avoid uncertainties connected with using incomplete models. Part I of the dismiltation reactionconvening a peroxide molecule into a water molecule with simultaneous oxidation of the metal atom of the catalystcan be done quite effectively at the Mn catalytic center. To act as catalytic scavengers of hydrogen peroxide, the oxomanganese salen complexes have to be deoxidized during part II of the dismutation reaction. It has been shown that there are two possible reaction routes for the second part of the dismutation reaction: the top and the side substrate approach routes. Our results suggest that the catalyst could be at least temporarily deactivated (poisoned) in the side approach reaction route due to the formation of a kinetically stable intermediate. Overall, the side approach reaction route for the catalyst recovery is the bottleneck for the whole dismutation process. On the basis of the detailed knowledge of the mode of action of the (salen)Mn-III catalase mimics, we suggest and rationalize structural changes of the catalyst that should lead to better therapeutic properties. The available experimental data support our conclusions. Our findings on the reaction dismutation mechanism could be the starting point for further improvement of salen-manganese complexes as synthetic catalytic scavengers of reactive oxygen species.