The U.S. Geological Survey, in cooperation with the City of Dallas Water Utilities Division, did a study to characterize bromide, chloride, and sulfate concentrations and loads at three U.S. Geological Survey streamflow-gaging stations on the reach of the Red River from Denison Dam, which impounds Lake Texoma, to the U.S. Highway 259 bridge near DeKalb, Texas. Bromide, chloride, and sulfate concentrations and loads were computed for streamflow-gaging stations on the study reach of the Red River. Continuous streamflow and specific conductance data and discrete samples for bromide, chloride, sulfate, and specific conductance were collected at three main-stem streamflow-gaging stations on the Red River: 07331600 Red River at Denison Dam near Denison, Texas (Denison Dam gage), 07335500 Red River at Arthur City, Texas (Arthur City gage), and 07336820 Red River near DeKalb, Texas (DeKalb gage). At each of these streamflow-gaging stations, discrete water-quality data were collected during January 2007-February 2009; continuous water-quality data were collected during March 2007-February 2009. Two periods of high flow resulted from floods during the study; floods during June-July 2007 resulted in elevated flow during June-September 2007 and smaller floods during March-April 2008 resulted in elevated flow during March-April 2008. Bromide, chloride, and sulfate concentrations in samples collected at the three gages decreased downstream. Median bromide concentrations ranged from 0.32 milligram per liter at the Denison Dam gage to 0.19 milligram per liter at the DeKalb gage. Median chloride concentrations ranged from 176 milligrams per liter at the Denison Dam gage to 108 milligrams per liter at the DeKalb gage, less than the 300-milligrams per liter secondary maximum contaminant level established by the Texas Commission on Environmental Quality. Median sulfate concentrations ranged from 213 milligrams per liter at the Denison Dam gage to 117 milligrams per liter at the DeKalb gage, also less than the 300-milligrams per liter secondary maximum contaminant level. Kruskal-Wallis analyses indicated statistically significant differences among bromide, chloride, and sulfate concentrations at the three gages. Regression equations to estimate bromide, chloride, and sulfate loads were developed for each of the three gages. The largest loads were estimated for a period of relatively large streamflow, June-September 2007, when about 50 percent of the load for the study period occurred at each gage. Adjusted R-squared values were largest for regression equations for the DeKalb gage, ranging from .957 for sulfate to .976 for chloride. Adjusted R-squared values for all regression equations developed to estimate loads of bromide, chloride, and sulfate at the three gages were .899 or larger.