Previous research has demonstrated that the reverse osmosis and advanced oxidation processes (AOPs) used to purify municipal wastewater to potable quality have difficulty removing low molecular weight halogenated disinfection byproducts (DBPs) and industrial chemicals. Because of the wide range of chemical characteristics of these DBPs, this study developed methods to predict their degradation within the AOP via UV direct photolysis and hydroxyl radical reaction, so that DBPs most likely to pass through the AOP could be predicted. Among 26 trihalomethanes, haloacetonitriles, haloacetaldehydes, halonitromethanes and haloacetamides, direct photolysis rate constants (254 nm) varied by ∼3 orders of magnitude, with rate constants increasing with Br and I substitution. Quantum yields varied little (0.12-0.59 mol/Einstein), such that rate constants were driven by the orders of magnitude variation in molar extinction coefficients. Quantum chemical calculations indicated a strong correlation between molar extinction coefficients and decreasing energy gaps between the highest occupied and lowest unoccupied orbitals (i.e., ELUMO-EHOMO). Rate constants for 37 trihalomethanes, haloacetonitriles, haloacetaldehydes, halonitromethanes, haloacetamides, and haloacetic acids with •OH measured by gamma radiolysis spanned 4 orders of magnitude. Based on these rate constants, a quantitative structure-reactivity relationship model (Group Contribution Method) was developed which predicted •OH rate constants for 5 additional halogenated compounds within a factor of 2. A kinetics model combining the molar extinction coefficients, quantum yields and •OH rate constants predicted experimental DBP loss in a lab-scale UV/H2O2 AOP well. Highlighting the difficulty associated with degrading these DBPs, at the 500-1000 mJ/cm2 UV fluence applied in potable reuse trains, 50% removal would be achieved generally only for compounds with several -Br or -I substituents, mostly due to higher molar extinction coefficients.