Despite the increased prominence of ecological measurement in fresh waters within recent national regulatory and legislative instruments, their assessment is still almost exclusively based on taxonomic structure. Integrated metrics of structure and function, though widely advocated, to date have not been incorporated into these bioassessment programmes. We sought to address this, by assessing community structure (macroinvertebrate assemblage composition) and ecosystem functioning (decomposition, primary production, and herbivory rates), in a series of replicated field experiments, at watercress farms on the headwaters of chalk streams, in southern England. The outfalls from watercress farms are typically of the highest chemical quality, however surveys have revealed long-term (30 years) impacts on key macroinvertebrate taxa, in particular the freshwater shrimp Gammarus pulex (L.), yet the ecosystem-level consequences remain unknown. Initial studies were at Europe’s largest watercress farm at St Mary Bourne, Hampshire, during the bioremediation of its complex wastewaters and changes to farm management practices. These widened to include larger scale spatiotemporal studies at other watercress farms. Detrimental ecological impacts at the start of the study were detected by the structural and functioning measures, but they did not respond to bioremediation. However, an increase in G. pulex abundance was detected, providing evidence of recovery in response to altered practices, which may be attributable to the cessation of chlorine use. The detrimental impacts were unique to the St Mary Bourne watercress farm and were not consistent across the other watercress farms in the study. Our results demonstrate the importance of integrated metrics of both ecosystem structure and functioning, to derive a more comprehensive view of aquatic ecosystems and highlights the difficulties associated with extrapolating from laboratory studies in response to stressors.