In February 2017, a Microcystis aeruginosa bloom (maximum 55 μg L−1 chl a and microcystin concentration of 40 μg L−1) occurred in a freshwater pond in Padre Island National Seashore (PINS), TX, USA, causing mortality of migratory redhead ducks and local animals. To treat the pond without further damage to the ecosystem, we proposed the use of H2O2. A nomograph of H2O2 required for oxidizing different initial cyanobacterial biomass levels to achieve desired final microcystin and chl a concentration was developed. With a chl a concentration > 50 μg L−1, the nomograph predicted > 97% removal efficiency when 50 mg L−1 peroxide was used. Then, an outdoor mesocosm experiment with four treatment levels (0, 10, 50 mg L−1 H2O2 and 10 mg L−1 H2O2 with ultrasonication) was tested. In the mesocosm experiment, discrete samples were assessed for cell morphology, pigment, and microcystin toxin content (n = 3) at six different time intervals during the 48-h treatment period. Cell morphology was affected after 4-h exposure at 50 mg L−1 H2O2 treatment. Overall, > 75% of microcystin and > 80% of chl a were destroyed after 48-h exposure to 50 mg L−1 H2O2. The pond was treated with 10 mg L−1 H2O2 which reduced chl a by 60% and microcystin by 35% after 2 days. Control of the cyanobacterial bloom using H2O2 was effective in lab and mesocosm experiments, but the cyanobacterial bloom recovered at PINS 2 weeks after the treatment. The recovery resulted from an insufficient H2O2 addition, identifying a need to re-evaluate current protocols. Overall, no residual H2O2 were detected in any scales of treatment.