Stream restorations that increase geomorphic stability can improve habitat quality, which should benefit selected species and local aquatic ecosystems. This assumption is often used to define primary restoration goals; yet, biological responses to restoration are rarely monitored or evaluated methodically. Macroinvertebrate communities were inventoried at 6 study reaches within 5 Catskill Mountain streams between 2002 and 2006 to characterize their responses to natural-channel-design (NCD) restoration. Although bank stability increased significantly at most restored reaches, analyses of variation showed that NCD restorations had no significant effect on 15 of 16 macroinvertebrate community metrics. Multidimensional scaling ordination indicated that communities from all reach types within a stream were much more similar to each other within any given year than they were in the same reaches across years or within any type of reach across streams. These findings indicate that source populations and watershed-scale factors were more important to macroinvertebrate community characteristics than were changes in channel geomorphology associated with NCD restoration. Furthermore, the response of macroinvertebrates to restoration cannot always be used to infer the response of other stream biota to restoration. Thus, a broad perspective is needed to characterize and evaluate the full range of effects that restoration can have on stream ecosystems.