The GPS (Global Positioning System) is a navigational tool available to the public, comprising in part a network of satellites in orbit broadcasting signals to GPS receivers on earth. Due in large part to the very accurate clocks in the system, a GPS receiver that receives a number of these signals simultaneously can then establish its location on earth. Some GPS receiver models have the capability to record a history of the latitude and longitude co-ordinates at known times, and these may later be downloaded and viewed on a computer using applicable software. Effectively, this provides the equivalent of a bread crumb trail of the path taken by the receiver, although the frequency of the data points can vary substantially from one model to another, and it may not be adjustable by the user. By plotting the information on a map, the trail can be viewed and the position-time history of the vehicle may become evident. The ability to obtain data from a GPS device that was present in a vehicle during a crash could therefore provide data that would interest an investigator. This paper examines the accuracy of the trail history recorded on several GPS devices of different brands and price levels, for the purpose of determining the accuracy to a known baseline trail established with an engineering-grade GPS data logging unit. Units available for approximately $100 up to one valued at over $1000 were used, with the latter unit designed for aviation use in addition to automotive use. Of particular interest was whether or not the recorded trail data could show subtle driver actions such as complete, or even partial, lane changes. For this research, the test vehicle was driven on a chosen course involving accelerating from a stop to highway speeds, steady cruising, and slowing to a stop. Various steering actions including lane changes were used. The resulting test data was compared. Subsequent testing compared the results from two units of the same brand and model.