Lumbar spine pathology accounts for billions of dollars in societal costs each year. Although the symptomatology of these conditions is relatively well understood, the mechanical changes in the spine are not. Previous direct measurements of lumbar spine mechanics have mostly been performed on cadavers. The methods for in vivo studies have included imaging, electrogoniometry, and motion capture. Few studies have directly measured in vivo lumbar spine kinematics with in-dwelling bone pins. This study tracked the in vivo three-dimensional motion of the entire lumbar spine (L1 to S1) in 10 healthy, young-adult subjects. Two 1.55 mm (0.062 in.) diameter Kirshner wires were inserted into each vertebra's spinous process under anesthesia. Motion capture cameras were used to track a triad of passive markers attached to the wires. Offsets between anatomical landmarks and tracking markers were established with a CT scan for each individual vertebra. Subjects were asked to perform various exercises including walking and voluntary range of motion. Subjects were able to complete all of the exercises. All subjects reported being adequately informed of all of the procedures and there were no neurological or orthopaedic complications. The range of the average inter-segmental range of motion was 4.26°-4.38° in the sagittal plane, 2.61°-4.00° in the coronal plane, and 4.11°-5.24° in the transverse plane. Using a direct (pin-based) in vivo measurement method, the motion of the human lumbar spine during gait was found to be triaxial. This appears to be the first three-dimensional motion analysis of the entire lumbar spine using indwelling pins. The results were similar to previously published data derived from a variety of experimental methods.