Osseoligamentous spinal specimens buckle under even a small vertical compressive force. To allow higher axial forces, a compressive follower load (FL) was suggested previously that approximates the curvature of the spine without inducing intervertebral rotation in both the frontal and the sagittal planes. In in vitro experiments and finite element analyses, the location of the FL path is subjected to estimation by the investigator. Such non-optimized FLs may induce bending and so far it is still unknown how this affects the results of the study and their comparability. A symmetrical finite element model of the lumbar spine was employed to simulate upright standing while applying a follower load. In analogy to in vitro experiments, the path of this FL was estimated seven times by different members of our institute's spine group. Additionally, an optimized FL path was determined and additional moments of +/-7.5Nm were applied to simulate flexion and extension. Application of the optimized 500N compressive FL causes only a marginal alteration of the curvature (cardan angle L1-S1 in sagittal plane <0.25 degrees). An individual estimation of the FL path, however, results in flexions of up to 10.0 degrees or extensions of up to 12.3 degrees. The resulting angles for the different non-optimized FL paths depend on the magnitude of the bending moment applied and whether a differential or an absolute measurement is taken. A preceding optimization of the location of the FL path would increase the comparability of different studies.