Traditional experimental methods are unable to study the kinematics of whole lumbar spine specimens under physiologic compressive preloads because the spine without active musculature buckles under just 120 N of vertical load. However, the lumbar spine can support a compressive load of physiologic magnitude (up to 1200 N) without collapsing if the load is applied along a follower load path. This study tested the hypothesis that the load-displacement response of the lumbar spine in flexion-extension is affected by the magnitude of the follower preload and the follower preload path. Twenty-one fresh human cadaveric lumbar spines were tested in flexion-extension under increasing compressive follower preload applied along two distinctly different optimized preload paths. The first (neutral) preload path was considered optimum if the specimen underwent the least angular change in its lordosis when the full range of preload (0-1200 N) was applied in its neutral posture. The second (flexed) preload path was optimized for an intermediate specimen posture between neutral and full flexion. A twofold increase in flexion stiffness occurred around the neutral posture as the preload was increased from 0 to 1200 N. The preload magnitude (400 N and larger) significantly affected the range of motion (ROM), with a 25% decrease at 1200 N preload applied along the neutral path. When the preload was applied along a path optimized for an intermediate forward-flexed posture, only a 15% decrease in ROM occurred at 1200 N. The results demonstrate that whole lumbar spine specimens can be subjected to compressive follower preloads of in vivo magnitudes while allowing physiologic mobility under flexion-extension moments. The optimized follower preload provides a method to simulate the resultant vector of the muscles that allow the spine to support physiologic compressive loads induced during flexion-extension activities.