The purpose of this study is to provide information regarding the stability of ephemeral streams on the north side of the Grand Valley, Colorado. The ungaged ephemeral streams in this semiarid region are of particular interest because (1) the underlying bedrock geology, Mancos Shale, is a sedimentary rock deposit that has been identified as a major contributor of salinity to the Colorado River and (2) despite infrequent flows of short duration, monsoon derived floods in these ephemeral streams can carry substantial amounts of sediment downstream, affecting up and downstream banks and channel cross sections. The study area is of interest as salinity, or the total dissolved solids concentration, in the Colorado River causes an estimated $300 to $400 million per year in economic damages in the United States and it is estimated that 62% of Upper Colorado River Basin dissolved-solid loads originate from geologic sources. In an effort to minimize salt contributions to the Colorado River from public lands administered by the Bureau of Land Management (BLM) a comprehensive three-pronged salinity control approach is being used which incorporates (1) controlling point sources of salinity; (2) controlling nonpoint sources of salinity; and (3) preventing nonpoint sources of salinity from persisting. In 2018, the U.S. Geological Survey, in cooperation with BLM, began an assessment of ephemeral streams located in the north side of the Grand Valley, Colorado, to characterize stream channel stability. The USGS developed a method for automatically extracting channel cross-section geometry from existing remotely sensed terrain models. Based on estimated flood stage and surrogate streamflows, hydraulic characteristics were calculated. Furthermore, the channel geometries and hydraulic characteristics were used to estimate channel stability utilizing a statistical model. In this ongoing study, cross-section stabilities were determined from a stream channel stability assessment for a subset of 1,406 visited locations out of a desired 13,415 cross sections which were delineated from remotely sensed terrain models. The application of Manning’s resistance equation in combination with multiple Logistic Regression models demonstrated that channel stability can be estimated with an 0.85 goodness of fit for a validation dataset when using a combination of drainage area, width to depth ratio, sinuosity, and shear stress as the explanatory variables. Stream channel stability was extrapolated for the remaining 13,415 unvisited cross sections using the multiple Logistic Regression model and defined explanatory variables. Mapping the ephemeral streams and their associated stabilities could be used to prioritize areas for BLM remediation or changes in management strategies to reduce sediment and salinity loading to the Colorado River.