The purpose of this study was to ascertain the correlation between diagnostic facet blocks and treatment outcome, both surgical and nonsurgical. One hundred twenty-six patients who had previously undergone diagnostic facet injections were reviewed. Eighty-two had subsequently undergone lumbar arthrodesis. The rest were treated with a variety of nonoperative modalities. Statistical analysis of accumulated data failed to show any significant correlation between the results of facet blocks and outcome of operative arthrodesis. In addition, statistical analysis failed to show any significant correlation between the facet block results and the outcome of nonoperative treatment. The authors concluded that lumbar facet joint injections cannot be used to determine appropriate patient treatment because they are not predictive of either surgical of nonsurgical success.