Abstract: Although the notion that streams are influenced by the character of their landscape at multiple spatial scales is not new, the relative degree to which local versus regional factors affect ecological function in streams is not fully understood, and can be different between geographically proximate watersheds. Anthropogenic disturbances to the landscape such as agricultural practices can be detrimental to stream ecosystems. This study examined the influences of local habitat and riparian corridor condition compared to regional landscape influences on benthic macroinvertebrate assemblages in a Midwestern agroecosytem. Twenty-four reaches in the North Fork and Upper Fork sub-basins of the Sugar Creek watershed, Wayne County, Ohio were sampled to better understand how different habitat and landscape factors affect the structure of macroinvertebrate assemblages in these impacted headwater streams. A total of 72,529 macroinvertebrates representing 79 families in 22 orders were collected during Autumn of 2005 and Spring 2006 to compare assemblage structure between watersheds and across seasons. Family richness, evenness, and diversity showed no difference attributable to watershed, while evenness and diversity exhibited seasonal differences. Chironomid abundance seemed to account for the seasonal change. Percent Ephemeroptera, Plecoptera, and Trichoptera (%EPT) was significantly influenced by watershed and season with the North Fork watershed exhibiting a higher abundance of these pollution tolerant and habitat sensitive taxa than the Upper Fork at all sample sites. A Geographic information system (GIS) was used to delineate sample watersheds and analyze landscape character. Proportion (%) of low Intensity residential, high intensity residential, industrial/commercial, deciduous forest, evergreen forest, mixed forest, row crop, pasture/hay, wooded wetland and herbaceous wetland were calculated per hydrologic unit. The dominant land uses in both study watersheds were crop, pasture, deciduous forest, and low intensity residential. The North Fork exhibited a significantly higher proportion of pasture and deciduous forest land types than the Upper Fork, which was dominated by row crops, then pasture and forest respectively. North Fork had almost twice the amount of deciduous forest as did the Upper Fork. Canonical correspondence analysis (CCA) was used to assess the macroinvertebrate family-environment relationship and variance partitioning determined the degree of influence of 8 local and 8 regional environmental factors on invertebrate assemblage structure in each study basin. Local habitat factors explained 25.8% of the total variance while regional landscape factors explained 23.6% of the total variance with 2.7% of the variability shared by both. Upper Fork sites were generally scattered along a silt/muck to cobble habitat gradient, while North Fork sites were arranged along a pasture-forest to rowcrop landscape gradient. The higher proportion of pasture and deciduous forest in the North Fork may explain the greater distribution of EPT taxa found in the watershed, while the greater proportion of crops and smaller proportion of forest in the Upper Fork may explain the greater influence of fine substrates in the watershed. Other environmental factors including glacial geology and groundwater influence may have also contributed to these differences by introducing coarser substrates and cooler, stream temperatures.