As the use of carbon dioxide (CO2)-enhanced oil recovery continues, understanding its impacts on the in situ oil well microbial communities is important. In this study, we update the already substantial investigation into five oil wells of the Olla and Nebo Hemphill fields in the southeastern United States with metagenomics. The Olla field has undergone CO2 injection, whereas the Nebo Hemphill field has not. Under anoxic conditions in Nebo Hemphill, our data suggest that methanogenic archaea are in competition with fermentative bacteria for shared substrates, leading to reduced biogenic methane production. However, in the Olla wells that had introduced CO2, we find genomic evidence for the growth and replication of bacteria able to respire oxygen. Based on the co-occurrence and potential replication of methanogens in these wells, we hypothesise that there are still anoxic niches for methanogens. Oxygen-utilising microbes may utilise phenolic substrates, creating a situation where more material is available for anaerobic growth, counterintuitively increasing the capacity for methane production in these wells. Our data suggest that this consortium may be the reason that greater methanogenesis is seen in these CO2-affected wells, and we hypothesise that the gas injection may have had contaminating air. Therefore, using existing wells for processes such as enhanced oil recovery or carbon capture and storage will not only depend on well history and in situ conditions, but also on the development of in situ microbial networks within modified wells.