Twelve sites on ephemeral to perennial streams within eight agricultural watersheds in central Iowa were studied over a 20 month period (1980-1982) to provide a description of basic ecological structure and function of stream ecosystems in intensively managed agricultural watersheds. In addition, assessments based on the following key ecological indices were made to test the River Continuum Concept relating to headwater streams: water quality, detrital decomposition rates, primary production, and invertebrate trophic organization. The drainages were differentially modified by intensive row crop development, grazing, removal of riparian vegetation, channelization, tile drainage, and sewage plant effluent;The most agriculturally modified headwater streams did not display ecological characteristics typical of headwater reaches studied in other parts of the country nor did they demonstrate all functional changes hypothesized by the RC-concept. Unlike "typical" headwater streams, these open canopy agricultural streams displayed impaired water quality (e.g. high nutrient and suspended solid concentrations), very warm summertime stream temperatures, extremely high primary production and respiration rates, slow detrital decomposition rates, and generalist collector-scraper invertebrate assemblages. In addition, the highly modified agricultural streams displayed very large summertime temperature and oxygen fluxes over diurnal periods, summertime production-respiration ratios < 1, and invertebrate communities with relatively low diversities but high densities and biomass. Less modified, tree-lined streams functioned much like headwater sites studied elsewhere;Certain watershed, nearstream, and instream management practices such as grassed waterways, riparian buffer area protection against cropping and grazing, suspension of maintenance channelization, and conservation tillage could be employed as low cost alternatives to lessen some of the water quality impairments and ecological impacts associated with intensive agricultural land use practices currently employed in central Iowa. However, large scale long term watershed manipulation experiments will be necessary to demonstrate or predict quantitatively the effects of such mitigation measures applied to agricultural watersheds.