The first and possibly only major rise of atmospheric oxygen, from pO2 ≤ 0.1% PAL (the present atmospheric level) to pO2 ≥ 10% PAL, appears to have occurred sometime before 2 Ga ago, although the exact time of and the cause(s) for the rise have been hotly debated. Equally important questions on the atmospheric oxygen concern its stability, especially the mechanisms regulating the atmospheric pO2 level and the causes and magnitude of pO2 variations since the first major rise of atmospheric oxygen. Previous efforts to model the pO2 variation during the Phanerozoic time have typically relied on secondary information, such as the carbon and sulfur isotopic records of sedimentary rocks, and on simple dynamics of the geochemical cycles of O, C, S, and P based on box-type models. As a result, many kinetic questions about the variation and stability of atmospheric oxygen could not have been answered. Here we quantitatively evaluate the dynamics and stability of atmospheric O2 and CO2, using recent experimental data, field observations, and a new model for the C-O coupled geochemical cycles. We examine the change with time in the fluxes of various compounds (O2, CO2, phosphate, organic C, carbonate C, C-bearing reduced volcanic gases, and C-free reduced volcanic gases) among the various reservoirs (atmosphere, soil, surface ocean, deep ocean, the lower crust and mantle, and upper crust) under a variety of scenarios. Our model does not assume steady-state fluxes for any of the reservoirs. Rather, the model incorporates the kinetic experimental data on oxidation of coal, a proxy for kerogen, the dynamics of soil formation and erosion, the kinetics of decomposition of organic matter in the Oceans by aerobic and anaerobic bacteria, the equilibrium ocean-atmosphere carbonate model, the observed relationships among the organic burial flux, dissolved O2 content of deep ocean, and sedimentation rates, and the three-box model ocean. The important parameters that strongly influence the dynamics of atmospheric O2, are found to be (a) the total area of soil formation on Earth; (b) the average soil depth; (c) the average rate of physical erosion of soils, which is linked to the average rate of accumulation of clastic sediments in the oceans; (d) the composition and flux of volcanic gas; and (e) the level of atmospheric CO2. We develop kinetic equations linking these parameters to the production and consumption fluxes of atmospheric oxygen and also to stable pO2 values. Considering the likely ranges of variations in these parameters in geologic history, we suggest that the atmospheric pO2 level is likely to have stayed within a very narrow range of 0.6-2 PAL and that the entire ocean, except for local euxinic basins, is likely to have been basically oxygenated since the first major rise of atmospheric oxygen more than 2 Ga ago.