Climate induced changes in runoff regimes and ongoing anthropogenic modification of land use and land cover (LULC) are shifting ambient water quality signals worldwide. Modulation of these signals by the physical catchment structure over different scales adds complexity to interpreting and analyzing measured data. Further bias may be introduced where monitoring networks are not representative of the structure of catchments in a given region. Here, we present a new environmental regionalization method to assess the representativeness of water quality monitoring (WQM) networks and to identify key structural drivers linked to water quality signals. Unique numerical codes were generated at the pixel level to provide wall-to-wall coverage of key Catchment Structural Units (CSUs) based on LULC, surficial geology, wetlands and slope. CSU codes were generated for all tributary (AT) catchments >20 km2 in Southern Alberta (n = 289), Canada, to determine the representativeness of an existing WQM network (54 tributary catchments) and to assess the explanatory power of CSUs with respect to water quality signals. Cluster analysis (CA) and multi-dimensional scaling (MDS) on the percent area of CSUs in the AT catchments identified six primary structural clusters in Southern Alberta. A clear gradient in catchment structure was evident progressing downstream from the Rocky Mountain headwaters through the foothills and prairie/plains region. Montane and grassland regions were found to be potentially under-represented by the current WQM program whereas catchments dominated by agriculture were likely over-represented. The disproportionate impact of specific CSU combinations on water quality was illustrated where the CA and MDS analyses indicated that even small percentages of urban areas and badland type topography results in elevated concentrations of total recoverable metals, nutrients and major ions. The application of the CSU approach in Southern Alberta demonstrates its value as an alternative method to assess and/or redesign existing WQM networks and to link water quality data to the structural composition of catchments. The general availability of the required data to generate CSUs provides universal potential for the approach to help assess other WQM programs and to contextualize data records. Applying the CSU approach when developing new ambient WQM networks can also help reduce the potential of over-monitoring similarly structured catchments as well as ensuring that all structural classes are represented by the data being generated.