Shock-tube experiments were performed in a mixture of 0.222% H-2/0.392% NO2/Ar between 1535 and 2003 K near 1.1 atm. Time histories of OH* chemiluminescence from the A -> X band near 307 nm were recorded and showed poor agreement with predictions from a recent hydrocarbon/NOx model when only the OH*-forming reactions N2O + H reversible arrow N-2 + OH* (R2) and O + H (+M) reversible arrow OH* (+M) (R3) were included. Since chemiluminescence is strongly correlated with heat release and since the reaction NO2 + H reversible arrow NO+OH is known to be primarily responsible for heat release during H-2-NO2 oxidation, the chemiluminescent reaction NO2 + H reversible arrow NO + OH* (R1) was proposed for the first time. By fitting the experimental OH* data, a best-fit rate constant was obtained as k(1) = 7.0 x 10(13) exp(-27,680/T), with k(1) in cm(3) mol(-1) s(-1) and T in K. This expression for k(1) is valid in the experimental temperature range of 1535 to 2003 K. The fitted k(1) value is dependent on the base NOx mechanism used. OH* profiles were also acquired in a mixture of 0.333% H-2/0.666% N2O/Ar between 1448 and 1776 K near 1.1 atm. The introduction of the new reaction R1 into the mechanism had no effect on the modeling of either the newly acquired H-2-N2O OH* data or previous H-2-N2O OH* data from the literature. Finally, R1 and R2 violate a long-held assumption concerning the exothermicity of such reactions, suggesting that the exothermicity criteria used to evaluate potential chemiluminescent reactions could be relaxed in future studies. Instead, a new methodology based on both the enthalpy of reaction and the entropy of reaction could be employed to identify new chemiluminescent reactions. To the best of the authors' knowledge, this is the first detailed study of OH* chemiluminescence kinetics in the H-2-NO2 system. (C) 2019 The Combustion Institute. Published by Elsevier Inc. All rights reserved.