In 1986, a mobility model was developed for predicting the traction and motion resistance of both wheeled and tracked vehicles on shallow snow, and a winter field season was dedicated to gathering mobility data for a diverse family of vehicles (including four on wheels and three tracked) to validate the model. The original version of the model, SSM1.0, used the Mohr-Coulomb shear failure equation from soil mechanics to predict gross traction. Motion resistance is predicted by calculating the amount of work done by the tire in compacting snow and only requires snow depth and density values as input snow properties. Some effort was expended in determining an easy and reliable method of obtaining snow strength parameters. The model was originally designed to use an initial snow density-snow strength relationship established from past instrumented vehicle test results, Historically, shear annulus apparati have been used to obtain Mohr-Coulomb strength parameters. A comparison of snow strength obtained via these three methods (shear annulus, instrumented vehicle, calculated from initial density using the relationship in SSM1.0) for individual snow covers showed no agreement. SSM1.0 assumed that snow strength parameters for mobility prediction were a function of initial snow density; however, traction is developed in the compacted snow under the driving element, whose strength properties bore little relation to those of the initial snow. It appears that the shear strength of the compacted snow is essentially a constant for all of the vehicles and snow covers tested here. Based on this finding, a new traction algorithm was developed, resulting in the creation of a second generation model, SSM2.0. This algorithm predicts gross traction, on the average for the vehicles tested, within 7% of the measured value. Motion resistance prediction remains unchanged in SSM2.0. This quantity is still not predicted with a desirable level of accuracy.