Studying aquatic biotransformation of chemicals in laboratory experiments, i.e., OECD 308 and OECD 309 studies, is required by international regulatory frameworks to prevent the release of persistent chemicals into natural water bodies. Here, we aimed to address several previously described shortcomings of OECD 308/309 studies regarding their variable outcomes and questionable environmental relevance by broadly testing and characterizing a modified biotransformation test system in which an aerated water column covers a thin sediment layer. Compared to standard OECD 308/309 studies, the modified system showed little inter-replicate variability, improved observability of biotransformation, and consistency with first-order biotransformation kinetics for the majority of 43 test compounds, including pharmaceuticals, pesticides, and artificial sweeteners. To elucidate the factors underlying the decreased inter-replicate variability compared to OECD 309 outcomes, we used multidimensional flow cytometry data and a machine learning-based cell type assignment pipeline to study cell densities and cell type diversities in the sediment and water compartments. Our here presented data on cell type composition in both water and sediment allows, for the first time, to study the behavior of microbial test communities throughout different biotransformation simulation studies. We found that sediment-associated microbial communities were generally more stable throughout the experiments and exhibited higher cell type diversity than the water column-associated communities. Consistently, our data indicate that aquatic biotransformation of chemicals can be most robustly studied in test systems providing a sufficient amount of sediment-borne biomass. While these findings favor OECD 308-type systems over OECD 309-type systems to study biotransformation at the water-sediment interface, our results suggest that the former should be modified toward lower sediment-water ratios to improve observability and interpretability of biotransformation.