Abstract Annual macroinvertebrate sampling was performed from 1994 to 2005 in the Mica Creek Experimental Watershed in northern Idaho to examine the relationship between contemporary timber harvest practices and the structure and diversity of aquatic macroinvertebrate assemblages. Four years of calibration, 4 years of postroad, and 4 years of postharvest (roads + harvest) data were used to analyze treatment effects. Metrics used to assess change included functional feeding group composition, organism densities, taxa richness indices, species diversity, and sediment tolerance indices. These assessments included changes to three macroinvertebrate taxa commonly used as bioindicators of stream habitat quality: Ephemeroptera (mayflies), Plecoptera (stoneflies), and Trichoptera (caddis flies) (EPT). In relation to pretreatment conditions, there were generally no major changes in functional feeding group composition or species diversity. Increases were observed in overall abundance and EPT abundance, but they largely can be explained by natural population variability. These results suggest that little or no change in macroinvertebrate communities resulted from road construction and timber harvest activities in the watershed. Substantial interannual variability was also observed, suggesting that infrequent biomonitoring may not provide adequate information to assess subtle disturbance effects in watersheds but can be an effective part of a comprehensive watershed monitoring program when coupled with other biotic and abiotic metrics.