To have confidence managing rivers under changing environmental pressures we must demonstrate thorough understanding of their response. Models express our understanding quantitatively. For 30 rivers in England, catchment attributes were combined with hydro-climatic time-series in hourly-resolution water quality model applications over a 15-year period. Retaining high-resolution input observations within simplified catchment representations makes geographically widespread application of process-based models achievable, whilst still representing diurnal cycles and quantifying ecosystem functioning. Background vulnerability assessments revealed eutrophic conditions (>30 µg chlorophyll-a L−1 as diatoms), oxygen stress (<5 mg DO L−1), and violation of safe pathogen levels (>9 CFU mL−1 of the faecal indicator organism, Escherichia coli) in 10, three, and 11 rivers respectively. Pathogen risk only considered treated effluent sources, not covering intermittent discharges or livestock contributions. By 2050, under a backdrop of uncertain change in climate, river quality is expected to worsen by 4.7 % for 10th percentile DO and 27.5 % for 90th percentile E. coli, with urban influence strongly determining sensitivity to change. Whilst deoxygenation vulnerability appears not widespread, faster future deteriorations are projected than elsewhere, such as in the USA. Eutrophication shows much spatio-temporal variability in change (with an average 5.1 % decrease in 90th percentile chlorophyll-a), seemingly controlled by local hydraulic factors and top-down biotic interactions. Across all indicators, riparian condition and channel hydrodynamics appear more important in controlling variability than regional differences in hydro-climatology. Assessment of comprehensive government-mandated interventions suggests partial offsetting of worsening DO, and further eutrophication decrease. E. coli deteriorations are effectively offset, although land management actions alone lead to further worsening.