Increased need to quantify adult insects emerging from streams as a part of foodweb and ecosystem studies has placed new demands on techniques used to sample adults. The population sampled must be better understood to establish the scope of inferences that may be drawn from emergence data. We used data from 2 different studies to compare the structure of insect assemblages represented by benthic samples and emergence-trap samples and to compare adult insect assemblages collected in emergence traps placed at mid-channel vs streambank locations. Nonmetric multidimensional scaling (NMDS) ordination showed that some components of the benthic assemblage were underrepresented in the emergence-trap samples and others were underrepresented in benthic samples. These differences were mainly caused by taxa that emerged by crawling out on the stream bank (e.g., Plecoptera) or whose larvae reside in habitats, such as stream margins, that are underrepresented with traditional benthic sampling (e.g., Dixidae). The flux of insects into traps placed mid-channel did not differ significantly from the flux into traps placed along the stream bank. Taxa collected by mid-channel and streambank traps overlapped considerably, but midchannel traps tended to collect proportionally more Trichoptera, Ephemeroptera, and Diptera, whereas streambank traps collected proportionally more Plecoptera. Our results can be used to improve trapping designs for future assessments of aquatic insect emergence in studies of insect behavior and life histories and as part of foodweb and ecosystem research.