The U.S. Geological Survey conducted a field study from July 2000 through June 2004, in cooperation with the Purgatoire River Water Conservancy District, Colorado Water Conservation Board, and Bureau of Reclamation, to characterize and quantify losses and gains in Picketwire, Baca, El Moro, Chilili, Enlarged Southside, Model, John Flood, and Hoehne irrigation canals. These canals divert streamflow from the Purgatoire River between Trinidad Dam and the city of Hoehne, Colorado. Discharge measurements were made along the eight canals during steady-state conditions to identify subreaches with losses or gains. Losses and gains were computed between main-channel measurement sites along each canal by equating inflows to outflows plus flow loss or gain in the subreach. As part of this study, multiple discharge measurements also were made at Picketwire, El Moro, Chilili, Enlarged Southside, Model, John Flood, and Hoehne canal headgates to compare standard Parshall flume-rated and measured discharge at the canal headgates. Results from the discharge measurements showed that Picketwire, Chilili, and Hoehne Canals generally lose flow from the headgate to the end of the canal, although some subreaches showed gains during some measurements. Losses in Picketwire Canal ranged from about 7 percent to about 23 percent of the headgate inflow, and Chilili Canal losses ranged from about 2 percent to about 34 percent of the headgate inflow. Hoehne Canal losses ranged from only about 2 to 7 percent of the headgate inflow, which is within the uncertainty of the measurements. El Moro Canal appears to lose flow in some subreaches and gain flow in other subreaches. Despite gains in some subreaches, measurements show flow losses of about 28 percent of the headgate inflow for the entire El Moro Canal. Losses and gains in Baca, Picketwire, Chilili, and Enlarged Southside canals may be affected by the length of time that the canal has been flowing. Losses in these canals appear to decrease the longer the canal has been continuously flowing. In some cases, subreaches of some of these canals go from losing to gaining flow. Unlike some of the other canals, losses and gains in El Moro and John Flood Canal do not appear to be related to how long the canal was flowing before the measurements were made. Losses and gains in El Moro Canal are probably related to the physical attributes of the canal, such as the canal construction and proximity to other canals. Field data indicate that El Moro Canal gains flow from and loses flow to other canals. Measurements made from the Model Canal headgate to Model Reservoir show canal losses and gains ranging from 1 to 5 percent of the headgate inflow, which is less than the uncertainty of the measurements. However, measured canal losses and gains from Model Canal downstream from Model Reservoir ranged from a loss of 59 percent to a gain of 1 percent of the subreach inflow. Measured discharges at the canal headgates were usually higher than the discharges determined using the standard Parshall flume discharge tables. Of the 102 discharge measurements made at the canal headgates, 72 of the measured discharges were higher than the corresponding discharges determined using the standard Parshall flume discharge tables. This means that about 70 percent of the time, the amount of flow that was diverted into the canals was underreported. All measured discharges at the Picketwire and El Moro headgates were higher than the corresponding flume-rated discharges, and all but one measured discharge at the Chilili headgate were higher than the corresponding flume-rated discharges. Discharges measured at the remaining headgates varied from 14 percent lower to 27 percent higher than the corresponding flume-rated discharges.