This report presents the results of a study by the Allegheny County Health Department (ACHD) and the U.S. Geological Survey (USGS) to determine the concentrations of fecal-indicator bacteria in the Allegheny, Monongahela, and Ohio Rivers (Three Rivers) in Allegheny County, Pittsburgh, Pa. Water-quality samples and river-discharge measurements were collected from July to September 2001 during dry- (72-hour dry antecedent period), mixed-, and wet-weather (48-hour dry antecedent period and at least 0.3 inch of rain in a 6-hour period) conditions at five sampling sites on the Three Rivers in Allegheny County. Water samples were collected weekly to establish baseline conditions and during successive days after three wet-weather events. Water samples were analyzed for fecal-indicator organisms including fecal-coliform (FC) bacteria, Escherichia coli (E. coli), and enterococci bacteria. Water samples were collected by the USGS and analyzed by the ACHD Laboratory. At each site, left-bank and right-bank surface-water samples were collected in addition to a composite sample (discharge-weighted sample representative of the channel cross section as a whole) at each site. Fecal-indicator bacteria reported in bank and composite samples were used to evaluate the distribution and mixing of bacteria-source streams in receiving waters such as the Three Rivers. Single-event concentrations of enterococci, E. coli, and FC during dry-weather events were greater than State and Federal water-quality standards (WQS) in 11, 28, and 28 percent of the samples, respectively; during mixed-weather events, concentrations of fecal-indicator bacteria were greater than WQS in 28, 37, and 43 percent of the samples, respectively; and during wet-weather events, concentrations of fecal-indicator bacteria were greater than WQS in 56, 71, and 81 percent of samples, respectively. Single-event, wet-weather concentrations exceeded those during dry-weather events for all sites except the Allegheny River at Oakmont. For this site, dilution during wet-weather events or the lack of source streams upgradient of the site may have caused this anomaly. Additionally, single-event concentrations of E. coli and FC frequently exceeded the WQS reported during wet-weather events. It is difficult to establish a short-term trend in fecal-indicator bacteria concentrations as a function of time after a wet-weather event due to factors including the spatial variability of sources contributing fecal material, dry-weather discharges, resuspension of bottom sediments, and flow augmentation from reservoirs. Relative to E. coli and enterococci, FC concentrations appeared to decrease with time, which may be attributed to the greater die-off rate for FC bacteria. Fecal-indicator bacteria concentrations at a site are dependent on the spatial distribution of point sources upstream of the station, the time-of-travel, rate of decay, and the degree of mixing and resuspension. Therefore, it is difficult to evaluate whether the left, right, and composite concentrations reported at a particular site are significantly different. To evaluate the significance of the fecal-indicator bacteria concentrations and turbidity reported in grab and composite samples during dry-, mixed-, and wet-weather events, data sets were evaluated using Wilcoxon rank sum tests. Tests were conducted using the fecal-indicator bacteria colonies and turbidity reported for each station for a given weather event. For example, fecal coliform counts reported in the left-bank sample were compared against the right-bank and composite samples, respectively, for the Ohio River at Sewickley site during dry-, mixed-, and wet-weather events. The statistical analyses suggest that, depending on the sampling site, the fecal-bacteria concentrations measured at selected locations vary spatially within a channel (left bank compared to right, right bank compared to composite). The most significant differences occurred between feca