Macroinvertebrate taxon tolerances to human disturbance have been key components of bioassessment since the early 20^th century. Initially, tolerances were used to evaluate organic pollution, but more recently, their use has been generalized to evaluate overall disturbance. Numeric tolerance values have been assigned to individual taxa in many previous studies, but their origins can be unknown, and the values can lack empirical basis. We used macroinvertebrate data collected at 1106 stream and river sites in 12 western states sampled during the Environmental Monitoring and Assessment Program (EMAP) in 2000 to 2004 to develop macroinvertebrate tolerance values for western US lotic ecosystems. We used principal components analysis to create a 9-variable synthetic disturbance variable (SDV) for overall human disturbance at those sites. The SDV included measures of nutrients, site-scale physical habitat, and catchment-scale land use. For taxa found at ≥20 sites, we calculated tolerance values at multiple levels of taxonomic resolution as the weighted (mean standard deviation) of the SDV at the sites with relative abundances as the weights. We then used the taxon tolerance values and relative abundances at each site to calculate an assemblage tolerance index (ATI) score at several levels of taxonomic resolution. These results should help inform discussions of the level of taxonomic resolution needed for bioassessments. We also discuss how the tolerance values could be used to improve the sensitivity of bioassessment metrics.