We present a kinetic study of the reactions of the OH radical with the molecules HCl and DCl over the temperature range 300–700 K in order to investigate the ‘intermediate’ region between room-temperature measurements and flame investigations. Ground-state OH radicals were generated by the repetitive pulsed irradiation of H2O vapour in a flow system, kinetically equivalent to a static system, and monitored in the time-resolved mode by resonance fluorescence at λ= 307 nm [OH(A2∑+→X2Π), (0, 0)] following optical excitation. Decay profiles of the OH radical in the presence of HCl and DCl were constructed following pre-trigger photomultiplier gating, photon counting, signal averaging and computerised analysis. A specially constructed high-temperature stainless-steel reactor, particularly designed for photon-counting measurements on a high-temperature system and critical to the extension of previous rate measurements of such reaction of OH by ca. 250 K, is described in detail with particular emphasis for procedures for minimising large background photon counts normally encountered in this type of system. The resulting rate data obtained in this investigation for the range 300–700 K may be summarised in the Arrhenius forms: OH (X2Π)+ HCl: kR=(2.94 ± 0.48)× 10–12 exp[–(446 ± 32)/T] cm3 molecule–1 s–1, OH (X2Π)+ DCl: kR=(4.04 ± 0.74)× 10–12 exp[–(718 ± 33)/T) cm3 molecule–1 s–1.These data are compared with the results of temperature-dependent studies that have been carried out by direct monitoring of the OH radical, employing both resonance absorption and resonance fluorescence of OH(A–X), on pulsed systems and discharge–flow systems up to temperatures of 460 K. The extrapolation of these results to flame temperatures, for which rate data for the reaction between OH + HCl have been reported from mass-spectrometric sampling of stable molecules, is considered in some detail. The temperature dependence of the diffusional loss of OH(X2Π), which dominates removal of the radical in the absence of HCl or DCl, when fitted to the standard form kdiff=C + n ln (T) yields n= 1.77 ± 0.14. Estimates of the geometrical parameters for light collection of the resonance fluorescence signal, coupled with the measured temperature dependence for diffusional loss, in turn yields D12(OH—He)= 0.28 ± 0.06 cm2 s–1 at s.t.p. The observed temperature dependence is further considered quantitatively in terms of the appropriate Lennard-Jones parameters for interaction.