Surface level ozone destruction in polar environments may be initiated by oxidation of bromide ions by ozone, ultimately leading to Br-2 production. Ab initio calculations are used to support the development of atmospheric chemistry models, but errors can occur in study of the bromide-ozone reaction due to inappropriate treatment of the many-electron species and the ionic nature of the reaction. In this work, a high level ab initio study is used to take into account the electronic correlation and the polarization effects. Our results show three possible pathways for the reaction. In particular, we find that this process, though endothermic on the singlet spin state surface, can be energetically feasible on the triplet surface. The triplet surface can be reached through photoexcitation of ozone or by the spin crossing of the potential energy surface. Because this process is known to occur in the dark, it may be that it occurs after intersystem crossing to a triplet surface. This paper also provides a starting point calibration for any future ab initio calculation studies of the bromide-ozone reaction, from the gas to the condensed phase.