Recent studies suggest that stream restoration at the reach scale may not increase stream biodiversity, raising concerns about the utility of this conservation practice. We examined whether reach-scale restoration in disturbed agricultural streams was associated with changes in macroinvertebrate community structure (total macroinvertebrate taxon richness, total macroinvertebrate density, Ephemeroptera, Plecoptera, Trichoptera [EPT] taxon richness, % abundance of EPT taxa) or secondary production (macroinvertebrate biomass over time). We collected macroinvertebrate samples over the course of 1 y from restored and unrestored reaches of 3 streams in southern Minnesota and used generalized least-square (GLS) models to assess whether measures of community structure were related to reach type, stream site, or sampling month. After accounting for effects of stream site and time, we found no significant difference in total taxon richness or % abundance of EPT taxa between restored and unrestored reaches. However, the number of EPT taxa and macroinvertebrate density were significantly higher in restored than in unrestored reaches. We compared secondary production estimates among study reaches based on 95^th-percentile confidence intervals generated via bootstrapping. In each study stream, secondary production was significantly (2–3×) higher in the restored than in the unrestored reach. Higher productivity in the restored reaches was largely a result of the disproportionate success of a few dominant, tolerant taxa. Our findings suggest that reach-scale restoration may have ecological effects that are not detected by measures of total taxon richness alone.