The City of Tulsa, Oklahoma, uses Lake Eucha and Spavinaw Lake in the Eucha-Spavinaw basin in northwestern Arkansas and northeastern Oklahoma for public water supply. Taste and odor problems in the water attributable to blue-green algae have increased in frequency. Changes in the algae community in the lakes may be attributable to increases in nutrient levels in the lakes, and in the waters feeding the lakes. The U.S. Geological Survey, in cooperation with the City of Tulsa, investigated and summarized nitrogen and phosphorus concentrations and provided estimates of nitrogen and phosphorus loads, yields, and flow-weighted concentrations in the Eucha-Spavinaw basin for three 3-year periods - 2002-2004, 2003-2005, and 2004-2006, to update a previous report that used data from water-quality samples for a 3-year period from January 2002 through December 2004. This report provides information needed to advance knowledge of the regional hydrologic system and understanding of hydrologic processes, and provides hydrologic data and results useful to multiple agencies for interstate agreements. Nitrogen and phosphorus concentrations were significantly greater in runoff samples than in base-flow samples for all three periods at Spavinaw Creek near Maysville, Arkansas; Spavinaw Creek near Colcord, Oklahoma, and Beaty Creek near Jay, Oklahoma. Runoff concentrations were not significantly greater than base-flow concentrations at Spavinaw Creek near Cherokee, Arkansas; and Spavinaw Creek near Sycamore, Oklahoma except for phosphorus during 2003-2005. Nitrogen concentrations in base-flow samples significantly increased downstream in Spavinaw Creek from the Maysville to Sycamore stations then significantly decreased from the Sycamore to the Colcord stations for all three periods. Nitrogen in base-flow samples from Beaty Creek was significantly less than in samples from Spavinaw Creek. Phosphorus concentrations in base-flow samples significantly increased from the Maysville to Cherokee stations in Spavinaw Creek for all three periods, probably because of a wastewater-treatment plant point source between those stations, and then significantly decreased downstream from the Cherokee to Colcord stations. Phosphorus in base-flow samples from Beaty Creek was significantly less than phosphorus in base-flow samples from Spavinaw Creek downstream from the Maysville station. Nitrogen concentrations in runoff samples were not significantly different among the stations on Spavinaw Creek for most of the three periods, except during 2003-2005 when runoff samples at the Colcord station were less than at the Sycamore station; however, the concentrations at Beaty Creek were significantly less than at all other stations. Phosphorus concentrations in runoff samples were not significantly different among the three downstream stations on Spavinaw Creek and were significantly different at the Maysville station on Spavinaw Creek and the Beaty Creek station, only during 2004-2006. Phosphorus and nitrogen concentrations in runoff samples from all stations generally increased with increasing streamflow. Estimated mean annual nitrogen total loads for the three 3-year periods were substantially greater at the Spavinaw Creek stations than at Beaty Creek and increased downstream from Maysville to Colcord in Spavinaw Creek, with the load at the Colcord station about 2 times that at Maysville station. Estimated mean annual nitrogen base-flow loads at the Spavinaw Creek stations were about 5 to 11 times greater than base-flow loads at Beaty Creek. The runoff component of the annual nitrogen total load for Beaty Creek was 85 to 89 percent; whereas, the range in the runoff component at the Spavinaw Creek stations was 60 to 71 percent. Estimated mean annual phosphorus total loads for the three 3-year periods were greater at the Spavinaw Creek stations from Cherokee to Colcord than at Beaty Creek and increased downstream from Maysville to Colcord in Spavinaw Creek, wit