The purpose of this report was to assess various methods of testing the mechanical properties of the spine and to suggest those most appropriate for particular applications. Questions addressed include the following: Are cadaveric experiments a reliable guide to mechanical properties in-vivo? Is it better to test "motion segments" or larger sections of spine? Is it necessary to simulate forces from individual muscles? How high should the applied forces be? The results of some recent experiments are described to indicate the potential of the techniques available.