Leaf litter decomposition is a key mechanism in headwater streams, allowing the transfer of nutrients and energy into the entire food web. However, chemical contamination resulting from human activity may exert a high pressure on the process, possibly threatening the structure of heterotrophic microbial communities and their decomposition abilities. In this study, the rates of microbial Alnus glutinosa (Alnus) leaf decay were assessed in six French watersheds displaying different land use (agricultural, urbanized, forested) and over four seasons (spring, summer, autumn, winter). In addition, for each watershed at each sampling time, both upstream (less-contaminated) and downstream (more-contaminated) sections were monitored. Toxicities (estimated as toxic units) predicted separately for pesticides and pharmaceuticals as well as environmental parameters (including nutrient levels) were related to microbial decay rates corrected for temperature and a range of fungal and bacterial community endpoints, including biomass, structure, and activity (extracellular ligninolytic and cellulolytic enzymatic activities). Results showed that agricultural and urbanized watersheds were more contaminated for nutrients and xenobiotics (higher pesticides and pharmaceuticals predicted toxicity) than forested watersheds. However, Alnus decay rates were higher in agricultural and urbanized watersheds, suggesting compensatory effects of nutrients over xenobiotics. Conversely, fungal biomass in leaves was 2-fold and 1.4-fold smaller in urbanized and agricultural watersheds than in the forested watersheds, respectively, which was mostly related to pesticide toxicity. However, no clear pattern was observed for extracellular enzymatic activities except that β-glucosidase activity positively correlated with Alnus decay rates. Together, these results highlight microbial communities being more efficient for leaf decomposition in polluted watersheds than in less contaminated ones, which is probably explained by changes in microbial community structure. Overall, our study showed that realistic chemical contamination in stream ecosystems may affect the biomass of Alnus-associated microbial communities but that these communities can adapt themselves to xenobiotics and maintain ecosystem functions.