Clean rivers, lakes, and streams consistently rank foremost among environmental priorities for citizens of Minnesota. Recent research has emphasized the importance of using physical, chemical, and biological indicators of stream health for diagnosing impaired watersheds and their receiving water bodies. A multidisciplinary team (14 members) of biologists and hydrologists and the Minnesota Pollution Control Agency (MPCA) is developing a regional stream classification system to facilitate Total Maximum Daily Load (TMDL) assessment of impaired waters in Minnesota. The classification will use regional reference reaches to account for complex factors related to precipitation, land use, soil, and geology. Objectives are: (1) investigate the relationship between indicators of stream health and local characteristics in 4 ecoregions, (2) develop and evaluate dimensionless curves using reference reach values for stream classification and TMDL assessment, and (3) test a stream classification system using the prediction intervals of the dimensionless curves. An independent data set will be used for validating the dimenionless curves. The approach is well suited to affect changes in land use and other watershed practices to address nonpoint source pollution. The classification system will be used by the MPCA to prioritize TMDL programs.