Knowledge on the behaviour of disinfection by-products (DBPs) during aquifer storage and recovery (ASR) is limited even though this can be an important consideration where recovered waters are used for potable purposes. A reclaimed water ASR trial in an anoxic aquifer in South Australia has provided some of the first quantitative information at field-scale on the fate and transport of trihalomethanes (THMs) and haloacetic acids (HAAs). The results revealed that THM half-lives varied from <1 to 65 days, with persistence of chloroform being highest and bromoform lowest. HAA attenuation was rapid (<1 day). Rates of THM attenuation were shown to be highly dependent on the geochemical environment as evidenced by the 2-5 fold reduction in half-lives at the ASR well which became methanogenic during the storage phase of the trial, as compared to an observation well situated 4 m away, which remained nitrate-reducing. These findings agree with previous laboratory-based studies which also show persistence declining with increased bromination of THMs and reducing redox conditions. Modelling suggests that the chlorinated injectant has sufficient residual chlorine and natural organic matter for substantial increases in THMs to occur within the aquifer, however this is masked in some of the field observations due to concurrent attenuation, particularly for the more rapidly attenuated brominated compounds. The model is based on data taken from water distribution systems and may not be representative for ASR since bromide and ammonia concentrations in the injected water and the possible role of organic carbon in the aquifer were not taken into consideration. During the storage phase DBP formation potentials were reduced as a result of the removal of precursor material despite an increase in the THM formation potential per unit weight of total organic carbon. This suggests that water quality improvements with respect to THMs and HAAs can be achieved through ASR in anoxic aquifers.