A complementary shock tube system was used to study dilute mixtures of carbon dioxide and hydrogen in a 1:5 ratio reacting behind reflected shock waves to produce water and carbon monoxide over the temperature range 2275-3860 K. One shock tube was outfitted to record simultaneously the infrared emissions from water and carbon dioxide through interference filters centered at 3.8 and 4.2 p, respectively. In order to minimize overlapping emissions, D2 was employed instead of H2. The formation of D2O exhibited quadratic dependence with respect to reaction time. The appearance of CO2 emission 4served primarily to mark time zero for the reaction and to establish the position of equilibrium. Argon was the inert diluent. The starting reactant percentages and the reflected shock zone pressure were varied for the purpose of determining the order dependencies for D2 and the total gas density. The second shock tube was connected to a time-of-flight mass spectrometer which recorded the histories of reactants, products, and intermediates. One notable feature of the TOF experiments was the detection of a small peak at m/e 29 corresponding to the HCO radical which was formed at times previous to significant CO production when H2 was used as a reactant. Runs performed on a, CO2-D2 mixture under similar conditions revealed the presence of DCO. The mole fraction for water formation from both shock tubes was fit to the expression (1 -/D2o//b2o,eq) * exp{-k [D2]0,3[M]0,7t2), where k = l02o o±o'2 exp(-81.4 ± 2.3/RT) cm3 mol-1 s-2. The units for the activation energy are kcal mol-1. Product profiles computed from a selection of literature rate constants making up an atomic mechanism did not compare favorably with the experimental results. The .possibility of a mechanism involving vibrational excitation of the reactants prior to transition state formation is discussed.