Inland waters are large contributors to global carbon dioxide (CO 2 ) emissions, in part due to the vulnerability of dissolved organic matter (DOM) to microbial decomposition and respiration to CO 2 during transport through aquatic systems. To assess the degree of this vulnerability, aquatic DOM is often incubated in standardized "biolability" assays. These assays isolate the dissolved fraction of aquatic OM by size filtration prior to incubation. We test whether this size selection has an impact on the bacterial community composition and the consequent dynamics of DOM degradation using three different filtering strategies: 0.2 μm (filtered-and-inoculated), 0.7 μm (generally the most common DOM filter size) and 106 μm ("unfiltered"). We found that bacterial community composition, based on 16S rRNA amplicon sequencing, was significantly affected by the different filter sizes. At the same time, filtering strategy also affected the DOM degradation dynamics. However, the dynamics of these two responses were decoupled, suggesting that filtration primarily influences biolability assays through bacterial abundance and the presence of their associated predators. By the end of the 41-day incubations all treatments tended to converge on a common total DOM biolability level, with the 0.7 μm filtered incubations reaching this point the quickest. These results suggest that assays to assess the total biolability of aquatic DOM should last long enough to remove filtration artefacts in the microbial population. Filtering strategy should also be taken into account when comparing results across biolability assays.