The lithium/lithium hydroxide technique, employing atomic absorption spectroscopy for atomic lithium determination, has been used to study the decay of hydrogen-atom concentrations in fuel-rich, atmospheric-pressure hydrogen/oxygen/diluent flames. The observed second-order rate constants for hydrogen-atom decay are interpreted in terms of the reactions, H+H+M→H2+M, Rate constant=kH, M H+OH+M→H2O+M, Rate constant=kOH,M where M represents any of the bulk species present in the burnt-gas region of the flame. Analyses of the dependence of the observed second-order rate constant on molecular hydrogen concentration in over one hundred and fifty flames containing as diluent, nitrogen, argon, helium, carbon dioxide or steam have led to the following values (in ml2 molecule−2 sec−1): kH,N2=1.0×10−32, kH,Ar=0.6×10−32, kH,He=1.3×10−32, kOH,N2=1.3×10−32, kOH,Ar=1.8×10−32, kOH,He=2.0×10−32, kOH,H2O=2.4×10−32, all at 1900° K. The rate constants kH,H2O and kOH,H2 cannot be determined individually because of the balanced reaction, H+H2OH2+OH, Equilibrium constant, K and the value (kH,H2O+KkOH,H2)=2.1×10−32 ml2 molecule−2 sec−1 at 1900° K, was obtained. The algebraic complexity of the rate of recombination in flames diluted with carbon dioxide (when partial reduction to carbon monoxide occurs) prohibits the determination of kH,CO, kH,CO2, kOH,CO, and kOH,CO2. A simplified procedure based on the assumption of equal thirdbody efficiencies for carbon dioxide and carbon monoxide leads to an effective rate constant for Reaction (1) of 1.4×10−32 ml2 molecule−2 sec−1 and an upper limit for the rate constant of Reaction (2) of 0.5×10−32 ml2 molecule−2 sec−1, with these species as third bodies.