This report provides data on dissolved-oxygen (DO) concentrations and identifies the environmental processes that most affect DO concentrations during base-flow periods in the lower reaches of Middle Fork and South Fork Beargrass Creek in Jefferson County, Kentucky. These reaches are affected by inputs from combined-sewer overflows. Sections of the lower reaches of the two streams run through single-family residential areas and public parks that are used extensively by local residents during the summer. Recreational fishing and wading also are common in the Middle Fork reach. Continuous-record data collected during the summer and early fall (July-September 1996 on the Middle Fork and July-October 1995 on the South Fork) at three monitoring sites along each reach indicate generally decreasing DO concentrations in the downstream direction except for the South Fork Beargrass Creek at Winter Avenue site where channel modifications have resulted in higher velocities along with shallower depths during low-flow conditions. The channel modifications at this site increased the reaeration-rate coefficient (a measure of the capacity of the stream to absorb oxygen through the air-water interface), increased the potential for algae to attach to the rough concrete surface, and increased algal exposure to sunlight. Synoptic data available for selected constituent concentrations were used to calibrate and verify a computer model (U.S. Environmental Protection Agency QUAL2E model) capable of simulating processes that affect DO concentrations in streams. The results of the study indicate that streamflow, reaeration, and sediment-oxygen demand (SOD) are the factors that most affect net production and depletion of DO in the lower reaches of Middle Fork and South Fork Beargrass Creek. For the QUAL2E model, streamflow is used in the determination of depth, which in tum is used to estimate the consumption of oxygen by SOD. Streamflow also is used in the determination of the reaeration-rate coefficient. From the QUAL2E simulations, DO concentrations (in the mass balance) attributed to reaeration and SOD were at least an order of magnitude greater than any of the other factors that can affect DO concentrations. Large diurnal variability in DO concentrations resulted at the monitoring sites located at upstream and downstream ends of the Middle Fork and South Fork reaches, but as indicated in model simulation, the net effect of photosynthesis and respiration on DO concentration was small. Nitrogen, ammonia, and carbonaceous biochemical-oxygen demand were present at low concentrations in each of the study reaches; the model results indicate these constituents did not have a substantial effect on DO concentrations. Model simulations indicated that lowering the SOD rate by 50 percent would result in a substantial improvement in DO concentrations in the Middle Fork Beargrass Creek reach for extremely low base-flow conditions but would result in only limited improvement in DO concentrations in the South Fork Beargrass Creek reach. However, no simulations for extremely low base-flow conditions were conducted for the South Fork Beargrass Creek reach. More information on SOD is needed for stream reaches affected by periodic inputs of effluent. In such stream systems, the temporal and spatial variability of SOD needs to be better defined.