This project proposes to study the effect of bracing on homogenous low back pain patients classified by the presence or absence of either excessive anterior or posterior translation on flexion-extension radiographs. Patients with back pain of four weeks' to one year duration are classified into four groups: retrodisplacement - worse in extension, retrodisplacement - worse in flexion, spondylolisthesis (of any type), and normal translation. Each classified group is randomly assigned to flexion, extension, or non-specific education treatment. Flexion and extension treatments include specific education and exercises as well as brace treatment. To accomplish these specific aims, the following hypotheses will be tested: (1) Patients with radiographic retrodisplacement (reverse spondylolisthesis) and low back pain of greater than four weeks' duration who demonstrate greatest displacement on an extension radiograph will respond more favorably to a flexion orthosis than three other patient groups: (a) those with the same x-ray findings treated with an extension orthosis, (b) those with the same x-ray findings treated with nonspecific low back education, and (c) patients with normal radiographic translation treated with the same disregard orthosis. (2) Patients with radiographic retrodisplacement and low back pain of greater than four weeks' duration who demonstrate greatest displacement on the flexion radiograph will respond more favorably to an extension orthosis than three other patient groups: (a) those with the same radiographic findings treated with a flexion orthosis, (b) those with the same radiographic findings treated with nonspecific low back education, and (c) patients with normal radiographic translation treated with the same extension orthosis. (3) Patients with radiographic anterodisplacement (spondylolisthesis of any type) and low back pain of four weeks' duration will respond more favorably to a flexion orthosis than three other patient groups: (a) those with the same x-ray findings treated with an extension orthosis, (b) those with the same x-ray findings treated with nonspecific low back education, and (c) patients with normal radiographic translation treated with the same flexion orthosis. (4) The response to either a flexion or extension orthosis can be predicted by: (a) radiographic measurements, and (b) clinical symptoms and signs, or psychosocial variables.