Developments in Acoustic Doppler Current Profiler (ADCP) technologies have made these instruments potentially useful for making measurements of discharge in rivers and large streams. Although there have been several laboratory studies and some field experiments, quantitative information on the performance of ADCP's under field conditions is relatively rare but essential to proper assessment of the potential uses and limitations of these instruments. This study was a comparative evaluation of river discharge data and ADCP data collected with conventional methods at 12 selected U.S. Geological Survey streamflow- gaging stations in the continental United States.

ADCP discharge measurements were made at the 12 sites in 1994. Twenty-six of the 31 measurements differed by less than 5 percent from the discharges determined with conventional methods. All 31 ADCP measurements were within 8 percent of the conventional method discharges.

The standard deviations of the ADCP measurements ranged from approximately 1 to 6 percent and were generally higher than the measurement errors predicted by error-propagation analysis of ADCP instrument performance. These error-prediction methods assume that the largest component of ADCP discharge measurement error is instrument related. The larger standard deviations indicate that substantial portions of measurement error may be attributable to sources unrelated to ADCP electronics or signal processing and are functions of the field environment.