Thermal effects, which accompany flash photolyses, are known to interfere with the determination of reaction rate constants. There are two approximate models currently being used in literature to estimate the magnitude of these effects (1,8). The first model (1) is the more widely accepted. It is based on the assumption that thermal effects are due to the cooling of reacting gas at the walls of the reaction vessel. The second model (8) is based on the assumption that thermal effects are due to nonuniformity in the concentrations of free radicals produced in flash photolysis; it neglects the heat exchange at the wall of the reaction vessel.It is shown that the second model can be used to calculate the magnitude of thermal effects in reaction vessels of reasonable length. The model was applied to calculate k,,,, the rate constant for the reaction 2Br + Br, + 2Br2. The value of k,,, is found to be very sensitive to the choice of model for thermal effects. At room temperature the most reasonable value of k,,,, using the second model, is (4.3 _+ 1.3) x 101° l2 mole-' s-'. This value agrees very well with independent determinations of k,,, using a stationary photochemical technique. The first model for treatment of thermal effects (1) was used previously to show that such effects do not influence the measured rates of chemical reactions, and calculations of rate constants using this model have not usually been attempted. In one case ( 3 , however, the first model (1) for thermal effects was employed to calculate a value for k~, , which was found to be six times larger than our value. Consequently, the second model (8) appears to be a better approximation for quantitative evaluation of thermal effects.Using the raw data (8) and kgr = 43 x 10' l2 mole-' S-l, the value of kAr, the recombination rate constant of Br atoms in excess of argon, was found to be (3.0 k 0.2) x lo9 l2 mole-' s-', which agrees well with data available in the literature.