The ecological classification of coastal waters has become an important issue in ecosystem water quality assessment. Previous studies have suggested that abiotic variables seem to be a suitable alternative to biological data for classifying coastal areas at different scales. The study presented here proposes a downscaling methodology for the classification of coastal waters at a regional scale within the NE Atlantic based on standardized data and objective decision rules. Physical variables (temperature, wave exposure, tidal range and radiation) were selected because of their ecological role, availability and statistical decision rules. This information was based on satellite data and mathematical modelling of natural coastal processes. The N and NW Spanish coastline was subdivided into 41 20-km segments that were classified according to physical variables using the self-organizing map and k-means algorithms. To validate the classification with biological data, 21 sites representing the entire range of physical typologies in the study area were simultaneously and consistently sampled. Intertidal macroalgae were identified in each of 10 quadrats of 50 × 50 cm for two to three transects per site, according to a stratified sampling procedure. The coverage of macroalgae was obtained by photographic analysis. The physical classification shows four typologies: Lower Rias, Upper Rias, West Cantabric and East Cantabric. Statistical analyses confirmed the ecological significance of these typologies at the tidal levels where seaweeds were the major structural element (lower and middle intertidal). According to the biological data, the greatest differences were found between the Upper Rias and the rest of the N and NW Iberian Peninsula coast. Thus, the classification methodology has potential application as a management tool.