The lumbar spine is a source of disability due to low back pain (LBP), yet the precise diagnosis is unknown in 80-90 percent of patients. The lifetime prevalence is 75 percent with a cost to the U.S. economy as high as 80 billion dollars. The problem is partly caused by mechanical overloading of the tissues and thus, there is some potential for both primary and secondary prevention. Biomechanical techniques have been effective in improving our understanding of the loading conditions leading to LBP, and in developing techniques for improved diagnosis and more effectual methods of treatment. Much progress has been made through the use of biomechanical models. Most models assume that the external moments are balanced by trunk musculature. Multiple muscle system models, employing agonist and antagonists, now are available to define 3D spine reaction forces. The static indeterminacy is taken care of either by simplification of the model or by linear or nonlinear optimization. Dynamic analysis has shown that vibrational and impact conditions (such as vehicle driving) can excite the natural frequency of the spine and lead to high spinal loadings. In vivo measurements have shown the resonant frequency of the lumbar spine to be 4-5 Hz and many vehicles excite those frequencies. New biomechanical techniques employing electromyography can estimate muscle load and muscle fatigue. Stereo photogrammetric techniques for establishing segmental kinematics have great potential for improving the diagnosis of spinal problems. These techniques are solidly based on prior in-vitro measurements of spinal kinematics. Mechanical fixation techniques, such as pedicle fixation, show great promise in improving the treatment of spinal problems. These have been extensively analyzed by both finite element techniques and in-vitro simulation so as to improve design as well as surgical technique.