The United States Geological Survey uses water-level (or stage) measurements to compute streamflow at over 8000 stream gaging stations located throughout the United States (waterwatch.usgs.gov, 2016). Streamflow (or discharge) is computed at five minute to hourly intervals from a relationship between water level and discharge that is uniquely determined for each station. The discharges are posted hourly to WaterWatch (waterwatch. usgs.gov) and are used by water managers to issue flood warnings and manage water supply and by other users of water information to make decisions. The accuracy of the water-level measurement is vital to the accuracy of the computed discharge. Because of the importance of water-level measurements, USGS has an accuracy policy of 0.02 ft or 0.2 percent of reading (whichever is larger) (Sauer and Turnipseed, 2010). Older technologies, such as float and shaft-encoder systems, bubbler systems and submersible pressure sensors, provide the needed accuracy but often require extensive construction to install and are prone to malfunctioning and damage from floating debris and sediment. No stilling wells or orifice lines need to be constructed for radar installations. During the last decade testing by the USGS Hydrologic Instrumentation Facility(HIF) found that radar water-level sensors can provide the needed accuracy for water-level measurements and because the sensor can be easily attached to bridges, reduce the construction required for installation. Additionally, the non-contact sensing of water level minimizes or eliminates damage and fouling from floating debris and sediment. This article is a brief summary of the testing efforts by the USGS HIF and field experiences with models of radar water-level sensors in streamflow measurement applications. Any use of trade names in this article is for descriptive purposes only and does not imply endorsement by the U.S. Government.