From 1970 to 2000, many types of total knee arthroplasties were introduced, including mobile bearing knees. In order to mechanically evaluate the designs, tests were developed for factors such as strength, laxity and constraint, kinematics, and wear. A number of the testing methods became ASTM or ISO standards, which are generally required for market introduction. In the year 2000, attention was given to testing methods specifically for mobile bearing knees, because of their mechanical differences from the more widely used fixed-bearing designs. The process began with the identification of potential problems unique to the mobile bearings themselves. These problems included bearing dislocation and deformation of the bearings due to stops and pivots. Wear testing methodology was also seen as important because of the original claim that mobile bearings would have reduced wear compared with fixed bearings due to the lower contact stresses. In this paper, the theoretical framework for defining the tests was specified. Specific tests were then defined for the following potential failure modes: dislocation ('spin-out'), dynamic dissociation ('spit-out'), deformation (due to stops, etc.), and endurance and deformation (due to overhang). Existing tests for total knee arthroplasty, in general, were also cited: cyclic fatigue of tibial trays, constraint (upper bearing surface), and wear (with a recommendation for differentiating upper and lower bearing wear). In the future, it is recommended that laboratories and manufacturers use these new tests to gain experience in their use and evaluate the safety of future mobile bearing designs.