The effect of urbanization on stream macroinvertebrate communities was examined by using data gathered during a 1999 reconnaissance of 14 sites in the Municipality of Anchorage, Alaska. Data collected included macroinvertebrate abundance, water chemistry, and trace elements in bed sediments. Macroinvertebrate relative-abundance data were edited and used in metric and index calculations. Population density was used as a surrogate for urbanization. Cluster analysis (unweighted-paired-grouping method) using arithmetic means of macroinvertebrate presence-absence data showed a well-defined separation between urbanized and nonurbanized sites as well as extracted sites that did not cleanly fall into either category. Water quality in Anchorage generally declined with increasing urbanization (population density). Of 59 variables examined, 31 correlated with urbanization. Local regression analysis extracted 11 variables that showed a significant impairment threshold response and 6 that showed a significant linear response. Significant biological variables for determining the impairment threshold in this study were the Margalef diversity index, Ephemeroptera-Plecoptera-Trichoptera taxa richness, and total taxa richness. Significant thresholds were observed in the water-chemistry variables conductivity, dissolved organic carbon, potassium, and total dissolved solids. Significant thresholds in trace elements in bed sediments included arsenic, iron, manganese, and lead. Results suggest that sites in Anchorage that have ratios of population density to road density greater than 70, storm-drain densities greater than 0.45 miles per square mile, road densities greater than 4 miles per square mile, or population densities greater than 125-150 persons per square mile may require further monitoring to determine if the stream has become impaired. This population density is far less than the 1,000 persons per square mile used by the U.S. Census Bureau to define an urban area.