Using density-functional-plane-wave-based and localized-orbital computational methods, we systematically examine the binding of molecular HCl at a variety of surface sites on crystalline nitric acid trihydrate (NAT), a step preceding the chlorine activation reactions that contribute to the depletion of stratospheric ozone at high latitudes. We pay particular attention to the role played by surface dangling (non-hydrogen-bonding) OH groups. After optimizing six low index faces, we find that NAT(001) and (001) faces are thermodynamically the most stable. Only one surface site on the (001) face, with one nearby dangling OH group, exhibits a high affinity for HCl. At this binding site, adsorbed HCl forms a strong H...O hydrogen bond with an NO3- ion and a weaker Cl...H hydrogen bond with a nearby H2O molecule. The interaction energy and enthalpy at 190 K corrected for zero-point energies are 23 and 25 kJ/mol, respectively. The presence of one strong binding site per simulation cell, versus at least three previously reported on the (0001) face of ice Ih (Mantz, Y. A.; Geiger, F. M.; Molina, L. T.; Molina, M. J.; Trout, B. L. J. Phys. Chem. A 2000, 105, 7037), leads to a prediction of a lower HCl surface coverage on NAT than on ice, qualitatively consistent with experiments conducted on these surfaces. Additionally, we present kinetic and thermodynamic evidence that molecular HCl, adsorbed near one or two dangling OH groups, does not dissociate on NAT. By contrast, molecularly adsorbed HCl likely dissociates when interacting strongly with two dangling OH group on the ice Ih (0001) face as reported in previously published theoretical studies (Svanberg, M.; Pettersson, J. B. C.; Bolton, K. J. Phys. Chem. A 2000, 104, 5787; Mantz, Y. A.; Geiger, F. M.; Molina, L. T.; Molina, M. J.; Trout, B. L. Chem. Phys. Lett. 2001, 348, 285).