Rapids are an integral part of bedrock‐controlled rivers, influencing aquatic ecology, geomorphology, and recreational value. Flow measurements in rapids and high‐gradient rivers are uncommon because of technical difficulties associated with positioning and operating sufficiently robust instruments. In the current study, detailed velocity, water surface, and bathymetric data were collected within rapids on the Colorado River in eastern Utah. With the water surface survey, it was found that shoreline‐based water surface surveys may misrepresent the water surface slope along the centerline of a rapid. Flow velocities were measured with an ADCP and an electronic pitot‐static tube. Integrating multiple measurements, the ADCP returned velocity data from the entire water column, even in sections of high water velocity. The maximum mean velocity measured with the ADCP was 3.7 m/s. The pitot‐static tube, while capable of only point measurements, quantified velocity 0.39 m below the surface. The maximum mean velocity measured with the pitot tube was 5.2 m/s, with instantaneous velocities up to 6.5 m/s. Analysis of the data showed that flow was subcritical throughout all measured rapids with a maximum measured Froude number of 0.7 in the largest measured rapids. Froude numbers were highest at the entrance of a given rapid, then decreased below the first breaking waves. In the absence of detailed bathymetric and velocity data, the Froude number in the fastest‐flowing section of a rapid was estimated from near‐surface velocity and depth soundings alone.