The kinetics of the chemically activated reaction between the ethyl radical and molecular oxygen are analyzed using quantum Rice-Ramsperger-Kassel (QRRK) theory for k(E) with both a master equation analysis and a modified strong-collision approach to account for collisional deactivation. Thermodynamic properties of species and transition states are determined by ab initio methods at the G2 and CBS-Q/B3LYP/6-31G(d,p) levels of theory and isodesmic reaction analysis. Rate coefficients for reactions of the energized adducts are obtained from canonical transition state theory. The reaction of C2H5 with O-2 forms an energized peroxy adduct with a calculated well depth of 35.3 kcal mol(-1) at the CBS-Q//B3LYP/6-31G(d,p) level of theory. The calculated (VTST) high-pressure limit bimolecular addition reaction rate constant for C2H5 + O-2 is 2.94 x 10(13)T(-0.44). Predictions of the chemically activated branching ratios using both collisional deactivation models are similar. All of the product formation pathways of ethyl radical with O-2, except the direct HO2 elimination from the CH3CH2OO. adduct, involve barriers that are above the energy of the reactants. As a result, formation of the stabilized CH3CH2OO. adduct is important at low to moderate temperatures; subsequent reactions of this adduct should be included in kinetic mechanisms. The temperature and pressure dependent rate coefficients for both the chemically activated reactions of the energized adducts and the thermally activated reactions of the stabilized adducts are assembled into a reaction mechanism. Comparisons of predictions using this mechanism to experiment demonstrate the necessity of including dissociation of the stabilized ethylperoxy adduct. Two channels are particularly important, direct HO2 elimination and reverse reaction to C2H5 + O-2, where the ratio of these rates is a function of temperature and pressure. The predictions, using unadjusted rate coefficients, are consistent with literature observations over extended temperature and pressure ranges. Comparison of a mechanism using 7 x 3 Chebyshev polynomials to represent k(T,P) to a conventional mechanism which used k(T) only (different values for k(T) at different pressures) showed good agreement. The kinetic implications for low-temperature ignition due to the direct formation of ethylene and HO2 from ethylperoxy are discussed.