This paper considers alternative methods for measuring the true ground speed of tractor vehicles as a means for controlling wheel slip to improve traction efficiency and reduce vehicle and tire wear. Test data are presented on the slip of both driven and undriven wheels for a tractor operating under various speed, load, and soil conditions. Data from a “fifth wheel” operating on smoothed and compacted soil behind a draft sled also is included. Vehicle speed determined from the speed of undriven wheels as well as vehicle speed derived from driven (slipping) wheels using a number of data processing strategies is not sufficiently accurate for either vehicle monitoring or closed loop control.The errors of single and dual beam doppler radar speed sensors caused by vehicle motion are analyzed. Single beam radar systems are shown to operate with an accuracy acceptable for operator control using signal integration times greater than approximately one second. The dual beam radar system exhibits better accuracy by eliminating the effect of quasi steady state offsets of vehicle pitch, roll, and yaw. Nevertheless, the rates of change of these angles still result in spreading the doppler return spectrum again requiring filtering or averaging to obtain an estimate of ground speed.