This study investigated the relationships of three major aquatic assemblages (diatom, macroinvertebrate, and fish) and environmental variables, including sub-basin, hydrology, land cover, and water quality variables on multiple scales. Samples were collected at 720 sampling sites on the Korean nationwide scale. Geological variables, including altitude and slope, showed a strong positive correlation with proportions of forest in land cover types and cobbles in substrates, while they were negatively correlated with water quality variables, including conductivity and total phosphorus. Considering the concordance of the different assemblages, species richness of fish and macroinvertebrates displayed significant correlation, and diatoms were significantly correlated with fish. However, diatoms did not show significant correlation with macroinvertebrates. Altitude and slope showed significant correlation with all biological variables of the three assemblages. Macroinvertebrates and fish showed positive relations with large substrate sizes. Indices of diatoms and macroinvertebrates well reflected the perturbation of water quality variables. However, fish indices showed a relatively low association with water quality variables, compared with those of diatoms and macroinvertebrates. These patterns were also confirmed by the ordination and prediction of biological indices with environmental variables through the learning process of a self-organizing map as well as random forest. Overall, our study supports the concept of multi-scale habitat filters and functional organization in streams, and is consistent with the recommended use of multiple biological indices with more than one assemblage for the assessment of the biotic integrity of aquatic ecosystems.