A bridge-scour study by the U.S. Geological Survey, in cooperation with the Nevada Department of Transportation, began in April 1996 to evaluate the Mesquite, Nevada, and Riverside, Nevada, bridges on the lower Virgin River using a sediment-transport model and historical geomorphic data. The BRIdge Stream Tube model for Alluvial River Simulation (BRI-STARS) was used to estimate bridge scour. The model was first calibrated using data for the Virgin River flood of March 12, 1995. Surveyed channel-geometry data were available at 11 cross sections for dates before and after the March 1995 flood to allow for evaluation of the model results. The model estimated the thalweg altitude within plus or minus 1 meter at 10 of the 11 cross sections. The calibrated model then was used to estimate the contraction, channel, pier, and total scour for synthesized hydrographs for 100- and 500-year floods at the two bridge sites. The estimated maximum total scour at the Mesquite bridge was 1.30 meters for the 100-year flood and 1.32 meters for the 500-year flood. The maximum total scour at the Riverside bridge was 1.90 meters for the 100-year flood and 2.01 meters for the 500-year flood. General scour was evaluated using stage-discharge relations at nearby streamflow-gaging stations, 1993-95 channel-geometry data, and channel-geometry data for the 100- and 500-year floods. On the basis of stage and discharge at the Littlefield, Arizona, gaging station, no long-term trend in aggradation or degradation was found. However, several cycles of aggradation and degradation had occurred during the period of record; the difference between the highest and lowest stage was 0.87 meter for a chosen low-flow discharge of 5.66 cubic meters per second for 1929-95. The value of 0.87 meter is probably the best estimate of general scour. The cross sections had an average scour depth of 0.07 meter between 1993 and 1994 and 0.16 meter between 1994 and 1995. The model simulated little general scour for the 100- and 500-year floods at the cross sections and did not give a good estimate of general scour, probably because the duration (days) of the floods used in the model was relatively short when compared with the duration (months or years) of geomorphic processes that influence long-term aggradation or degradation. Historical geomorphic changes of the Virgin River at the bridge sites and the causes of those changes were documented using aerial photographs from 1938-95 and other historical information. The Virgin River has become narrower and more sinuous through time, the vegetation on the flood plain has increased, and the channel has shifted laterally many times. The processes associated with these channel changes were found to be long-term changes in precipitation and streamflow; the duration, magnitude, and timing of floods; sediment-transport characteristics; channel avulsion; changes in density of vegetation; and anthropogenic influences.