From January 1980 through December 1984, 454 patients were evaluated with facet joint injections. All had the chief complaint of low-back pain, normal neurologic examinations and no root tension signs. Three hundred and ninety completed the protocol, which included a lumbar motion pain assessment before and after facet injection. A total of 127 variables were studied. There were 229 males and 161 females with a median age of 38. Facet joint arthrograms were performed prior to intra-articular injection of local anesthetic and cortisone. Initial mean pain relief was only 29%. Variables correlating significantly (P less than 0.05) with more postinjection pain relief were older age, prior history of low-back pain, normal gait, maximum pain on extension following forward flexion in the standing position, and the absence of leg pain, muscle spasm and aggravation of pain on Valsalva. Greatest pain relief immediately after injection was seen with lumbar extension and rotation, motions reported to stress the facet joints or aggravate pain of facet joint origin. Patients with more pain on lumbar extension and rotation as a group, however, did not get more pain relief. From this study we were not able to identify clinical facet joint syndromes or predict patients responding better to this procedure. The facet joints were not commonly the single or primary source for low-back pain in the great majority (greater than 90%) of patients studied.