The hot springs of Hot Springs National Park consist of a mixture of water from two recharge components: a primary hot-water component and a secondary cold-water component. Widespread distribution of fractures enables mixing of the hot- and cold-water components of flow near the discharge area for the springs. Urbanization in the area near the hot springs of Hot Springs National Park has increased the potential for degradation of the quality of surface-water runoff and locally derived ground-water recharge to the hot springs. Previous studies by the U.S. Geological Survey have indicated that water from some cold-water springs and wells in the vicinity of Hot Springs, Arkansas, showed evidence of contamination and that water from locally derived cold-water recharge might contribute 25 percent of the total flow to the hot springs after storms. Water samples were collected during base-flow conditions at nine hot springs and two cold-water springs in September 2000. Nine hot springs and one cold-water spring were resampled in October 2001 after a storm that resulted in a measurable decrease in water temperature in selected hot springs. Water samples were analyzed for a variety of dissolved chemical constituents (nutrients, major ions, trace elements, pesticides, semivolatile compounds, isotopes, and radiochemicals), physical properties, field measurements, and bacteria. Comparison of analyses of samples collected during base-flow conditions from the springs in 2000 and during a storm event in 2001 with the results from earlier studies dating back to the late 1800's indicates that little change in major, minor, and trace constituent chemistry has occurred and that the water continues to be of excellent quality. Water-quality data show distinguishable differences in water chemistry of the springs during base-flow and stormflow conditions, indicating changing input of cold-water recharge relative to hot-water recharge. Silica, total dissolved solids, strontium, barium, and sulfate show statistically significant differences between the median values of base-flow and stormflow samples. While variations in these constituents do not degrade water quality, the differences do provide evidence of variability in the factors controlling water quality of the hot springs and show that water quality is influenced by the locally derived, cold-water component of flow to the springs. Water temperature was measured continuously (3-minute intervals) between August 2000 and October 2002 at four hot springs. Continuous water-temperature data at the springs provide no indication of persistent long-term change in water temperature through time. Short time-scale water-temperature decreases occur in response to mixing of hot-springs water with locally derived recharge after storm events; the magnitude of these decreases varied inversely with the amount of rainfall. Maximum decreases in water temperature for specific storms had a non-linear relation with the amount of precipitation measured for the events. Response time for water temperature to begin decreasing from baseline temperature as a result of storm recharge was highly variable. Some springs began decreasing from baseline temperature as quickly as 1 hour after the beginning of a storm; one spring had an 8-hour minimum response time to show a storm-related temperature decrease. Water-quality, water-temperature, isotopic, and radiochemical data provide multiple lines of evidence supporting the importance of the contribution of cold-water recharge to hot springs. All the springs sampled indicated some measure of influence from local recharge. Binary mixing models using silica and total dissolved solids indicate that cold-water recharge from stormflow contributes an estimated 10 to 31 percent of the flow of hot springs. Models using water temperature indicate that cold-water recharge from stormflow contributes an estimated 1 to 35 percent of the flow of the various hot springs. Alth