Bioassessment is widely used to measure ecological integrity of natural habitats following anthropogenic dis-turbances and modifications. Traditionally, bioassessment has been based exclusively on species-environment interactions, i.e. niche processes. However, dispersal processes, and in particular mass effect, could mask the influence of niche processes and lead to erroneous conclusions about ecosystem health. To circumvent this problem, we identified 40 diatom species with distributions driven primarily by mass effect and propose an alternative version of the Biological Diatom Index (IBD2007) excluding these species. We tested the environmental responses of both the original IBD (IBD2007) and the modified IBD (IBDmod) with a benthic diatom dataset from France, collected between 2007 and 2013 and including 9487 samples from 3913 spatially distinct localities. Our results indicate a better relationship between the IBDmod scores and environmental conditions, compared to the IBD2007 scores, leading to a more accurate determination of river ecological status, especially in conditions of moderate nutrient enrichment. This study supports the idea that mass effect may result in biased evaluation of water quality. It is advocated that this process is considered in other diatom-based indices, and by extension, in any biotic index.