Following decades of riparian buffer zone (RBZ) studies there remains a need to look across individual site data for collective evidence on the site-specific pollution mitigation and river water quality. We explored primary study evidence on runoff, sediment, P, N, coliforms and pesticides using complimentary styles of metadata interpretation. A quantitative assessment of pollution retention (75 studies, 474 data rows) derived relationships for retention versus width, including significant covariates of clay particle size and buffer slope for sediment, total and dissolved P. Total N and coliforms related to texture and slope but were independent of width. Other factors across pollutants were inconsistently reported. With limitations on quantitative studies a second approach examined factor significance (formal testing versus inferred; 93 studies) across source pressure, transport/physical, vegetation and soil biogeochemical factors on pollution effectiveness. The RBZ evidence showed considerable disagreement and bias in shorter term study implications on longer-term processes. Screening for stronger evidence by study number and agreement left fifteen factors informing on at least one pollutant, whereas only rainfall intensity, preferential deposition, tree planting, soil infiltration remained addressing three or more pollutants. Key messages were that: data complexities, from short-term trapping in upper buffer edges and so-called ‘negative effectiveness’ associated with internal recycling and/or errors in constraining mass inputs for dissolved pollutant and subsurface transport require careful interpretation; RBZ intervention and study durations were limited compared to effect times (particularly vegetation management and changing soil conditions); factors affect pollutants with particulate and dissolved phases differently and must be understood to limit RBZ pollution swapping. Buffer functioning is highly site-specific. To understand this better attention should be given to revisiting studies of vegetation management to extend timeframes, wider study of belowground (soil biogeochemical and transport) processes and studies should document site contexts across source pressures, riparian hydrological, soil and vegetation factors.