Red blood cells (RBC) are considered as a circulating sink of H2O2, but a significant debate remains over the role of the different intraerythocyte peroxidases. Herein we examined the kinetic of decomposition of exogenous H2O2 by human RBC at different cell densities, using fluorescent and oxymetric methods, contrasting the results against a mathematical model. Fluorescent measurements as well as oxygen production experiments showed that catalase was responsible for most of the decomposition of H2O2 at cell densities suitable for both experimental settings (0.1–10 × 1010 cell L−1), since sodium azide but not N-ethylmaleimide (NEM) inhibited H2O2 consumption. Oxygen production decreased at high cell densities until none was detected above 1.1 × 1012 cell L−1, being recovered after inhibition of the thiol dependent systems by NEM. This result underlined that the consumption of H2O2 by catalase prevail at RBC densities regularly used for research, while the thiol dependent systems predominate when the cell density increases, approaching the normal number in blood (5 × 1012 cell L−1). The mathematical model successfully reproduced experimental results and at low cell number it showed a time sequence involving Prx as the first line of defense, followed by catalase, with a minor role by Gpx. The turning points were given by the total consumption of reduced Prx in first place and reduced GSH after that. However, Prx alone was able to account for the added H2O2 (50 µM) at physiological RBC density, calling attention to the importance of cell density in defining the pathway of H2O2 consumption and offering an explanation to current apparently conflicting results in the literature.