The kinetics for the uptake of N 2 O 5 with NaCl [N 2 O 5 (g) + NaCl(s) f ClNO 2 (g) + NaNO 3 (s) (reaction 1)] and KBr[N 2 O 5 (g) + KBr(s) f products (reaction 6)] have been studied in a Teflon-coated Knudsen reactor. The product of reaction 1 is found to be ClNO 2 , in agreement with previous studies. The only brominecontaining gaseous product observed for reaction 6 is Br 2 ; we propose a redox reaction in which Br -is oxidized to molecular bromine with the concurrent formation of nitrite. The hypothesis is supported by the observation of nitrous acid in the product spectrum of reaction 6. The observed uptake coefficients are found to depend strongly on the total external surface area of the salt substrates, prepared by a number of methods. With samples of well-defined total external surface, we are able to determine the following values for the uptake coefficient: γ 1 ) (5.0 ( 2.0) × 10 -4 ; γ 6 ) (4.0 ( 2.0) × 10 -3 . The values are smaller than previously reported. The unexpected dependence of the uptake kinetics on the surface presentation led us to conduct further experiments on the HNO 3 reaction with salt: HNO 3 (g) + NaCl(s) f HCl(g) + NaNO 3 (s) (reaction 2). From our new experiments conducted on various substrates, we find γ 2 ) (2.0 ( 1.0) × 10 -2 , in good agreement with our previous measurements on salt powder. The validity of calculating correction factors for the uptake coefficients, as recently proposed by Keyser and co-workers, is discussed in detail, and specific experimental tests of the theory are described. The role of internal surfaces in the overall observed uptake kinetics for the reactions of N 2 O 5 is found to be remarkably well described by the treatment given by Keyser and co-workers. Additional experiments on the uptake kinetics of nitric acid, however, cast doubt that this treatment can be generally extended to all species of atmospheric interest. The atmospheric implications of these findings are discussed briefly.