We established a pattern of uncultivated areas (corridors and a simulated roadside) within a soybean agroecosystem to determine how such areas would affect arthropod densities, levels of crop damage, and crop yield. Twelve 0.4-ha soybean plots were located along a simulated grassy roadside that we hypothesized would serve as a major reservoir for arthropods. Six of these plots were subdivided into thirds by 2 m wide uncultivated corridors either planted in grasses or allowed to enter secondary succession. Undivided crop plots were treated with insecticide (Sevin) or left untreated as controls. Effects of this spatial patterning were evaluated in 1980, 1 yr after establishment.Yield (soybeans per metre of soybean row) differed significantly among all treatments, with the yield in insecticide treatments > treatments divided by grassy corridors > undivided controls > treatments divided by successional corridors. Crop leaf damage reached economically injurious levels in all but insecticide-sprayed treatments. Damaged leaves displayed the lacework pattern characteristic of feeding by the Mexican bean beetle (Epilachna varivestis), whose population densities increased dramatically late in the growing season.Plots divided by successional corridors, which had the lowest yields and the highest levels of defoliation, also had significantly higher abundances of Epilachna varivestis, lepidopteran larvae, leafhoppers, and murids than controls during at least one sampling period. Levels of leaf damage and arthropod abundances generally were similar between control plots and plots divided by grassy corridors. Plots divided by grassy corridors, however, did contain fewer adult Empoasca fabae early in the growing season indicating possible impact of grassy areas on the distribution of adults colonizing the crop. E. fabae avoided grassy corridors and was significantly less abundant in the soybeans adjacent to grassy corridors, suggesting that grassy areas confer an "associational resistance" to invasion by migrating adults.Nomuraea rileyi, a fungal pathogen, infested significantly higher proportions of the green cloverworm (Plathypena scabra) in plots divided by corridors, especially in plots divided by grassy corridors. This fungus is the primary natural biotic control agent of this major lepidopteran larva pest in the Midwest. Presence of uncultivated areas may increase the rate of development ofepizootics of N. rileyi. Green cloverworm populations, however, were not significantly lower in divided plots despite these higher levels of fungal infection.Predators were more abundant in uncultivated areas, especially successional corridors, than in the soybean crop. However, uncultivated areas did not "funnel" predators into the crop as hypothesized. In fact, results suggest that some predators (e.g., Orius insidiosus) were attracted to successional corridors from the adjacent soybean crop.Findings suggest that management of uncultivated corridors within the landscape represents a potentially important means of regulating insect pest...