The loads on internal spinal fixation devices were measured using modified, telemeterized AO-Dick internal fixators. The implants allow measurement of the three force components and the three moments acting on the implant. The modifed fixators were mounted on cadaver spines, and the implant loads were measured in the intact and postcorpectomy spines for different loading modes, including axial compression force, flexion, extension, lateral bending, and torsion. The in vitro experiment did not consider muscle forces. Modified fixators were also implanted in three patients, and the implant loads were determined before and after anterior interbody fusion with autologous iliac-crest bone grafts. The results for different in vitro loading modes were compared with those in vivo in order to demonstrate the extent to which the in vitro loads represent the real situation in patients. In several cases, the implant loads in the in vitro experiment differed strongly from those measured in patients. For flexion and lateral bending, a tensile axial force occasionally was measured in the in vitro experiment, while in the patients the axial force was always compressive. Extension was predominantly associated with extension bending moments in the in vitro study but with flexion bending moments in the patients. When muscle forces are not considered in the in vitro experiment, the loads on the fixators may differ significantly from the situation found in patients.