The Arkansas River is an important municipal water supply and is the primary supply for about 400,000 acres of irrigated land in southeastern Colorado. The suitability of this water for domestic, agricultural, and industrial use is affected by high salinity in parts of the Arkansas River. There is a need to quantify mass loading of dissolved solids (DS) in the Arkansas River. In 2009, the U.S. Geological Survey, in cooperation with the Arkansas River Basin Regional Resource Planning Group and the Colorado Water Conservation Board, began a study to estimate gains and losses from unmeasured sources and sinks for streamflow and DS load in selected reaches of the Arkansas River in southeastern Colorado. Two study reaches were selected for investigation—Canon City to just upstream from Pueblo Reservoir (UARB) and Avondale to Las Animas (LARB). The results from the water-budget analyses indicated that potential areas of unmeasured sources and sinks of streamflow were identifiable in the two study reaches. In the UARB, a substantial volume of water in the subreach from Ark at Canon City to the seasonal gaging station 5 miles downstream (Ark nr Canon City) was unaccounted for by the methodology used in this analysis. The daily gain from unmeasured sources in this subreach was estimated to be about 100 cubic feet per second (ft³/s) or about 20 ft³/s per river mile. Water-budget estimates for the remaining 18 miles of the UARB study reach indicated that gains or losses from unmeasured sources or sinks were within the measurement error as defined for this report. In the LARB, gains and losses from unmeasured sources and sinks were identified in some of the subreaches but the magnitude of the flux generally was small. Unmeasured sources ranging from less than 2 to 3 ft³/s per mile were identified in the river subreaches from Ark at Catlin Dam downstream to Ark at Swink. A streamflow loss was indicated along the subreach from Ark at Nepesta to Ark at Catlin Dam, particularly in 2010. The mechanism and spatial extent of this sink was not identified, and further investigation would be required to better quantify the loss. The results from the analyses of unmeasured sources of DS load indicated that potential source areas were identifiable in the study areas. It might be expected that unmeasured DS load flux would be identified along the same reaches where unmeasured streamflow flux was identified. To that extent, some of the observed results from the analysis of daily DS loading did mirror the streamflow results. In some subreaches of the Arkansas River, however, unmeasured sources and sinks of DS load did not appear to be directly associated with unmeasured sources and sinks of streamflow. In the UARB from Ark at Canon City to Ark nr Canon City, unmeasured gains in DS load were estimated to range from 11 to 22 tons per day per mile in 2009 and from about 8 to 13 tons per day per mile in 2010; streamflow from unmeasured sources was estimated to be about 20 ft³/s per mile along this same reach. Downstream from this short reach, DS load to the river from unmeasured sources was estimated to range from 5.4 to 7.6 tons per day per mile in 2010 for Ark nr Canon City to Ark at Portland and from 11 to 16 tons per day per mile in 2009 for Ark at Portland to Ark nr Portland. Unmeasured gains in streamflow were not identified in either of these subreaches. Several small tributaries with DS concentrations ranging from 3,000 mg/L to as high as 6,000 mg/L enter the river along these subreaches. These inputs may indicate a potential source of groundwater that could affect DS loading in the river. Further investigation would be needed to identify the unmeasured source or sources of DS load to determine the nature and extent of unmeasured inputs. In the LARB, gains in DS load from unmeasured sources were identified for the subreach from Ark nr Avondale to Ark at Nepesta, although no substantial amounts of streamflow from unmeasured sources were identified for this subreach. In 2009, the estimated gain in DS load from unmeasured sources for this subreach was 4.7 tons per day per mile. An increase in DS load from unmeasured sources also was identified along the subreach of the river from Ark at Catlin to Swink; the DS load from unmeasured sources was estimated to range from 10 to 28 tons per day per mile. The only loss of DS load was identified for the subreach from Nepesta to Catlin Dam in 2010. The mechanism and spatial extent of the losses were not identified, and further investigation would be required to better understand the results.