Epidural steroid injection (ESI) is one of the most common nonsurgical interventions prescribed for back and leg pain symptoms. Although the use of ESI is widespread, proof of efficacy among the broad population of low back pain patients is lacking and use is predicated to a great extent on the cost and morbidity of the perceived "next step" in many patient's care-surgery. To review the relative indications and clinical features that predict success with ESI therapy, and to provide a physiological rationale to guide clinical decision-making. Review of literature and clinical experience. Clinical studies have alternately supported and refuted the efficacy of ESI in the treatment of patients with back and leg pain. Steroid medications do benefit some patients with radicular pain, but the benefit is often limited in duration, making efficacy difficult to prove over time. Steroids appear to speed the rate of recovery and return to function, however, allowing patients to reduce medication levels and increase activity while awaiting the natural improvement expected in most spinal disorders. Fluoroscopic verification of needle placement, with contrast injection, greatly improves steroid delivery while reducing risks. Although it is assumed that the benefit of steroids is related to their effect on inflammation, that remains unproven, and it is possible that benefit is gained through an unrecognized action. Randomized, controlled trials are needed to conclusively identify those patients most likely to benefit from ESI, and when and for how long. Until then, epidural steroids provide a reasonable alternative to surgical intervention in selected patients with back and/or leg pain, whose symptoms are functionally limiting. When appropriate goals are established and proper patients are selected, sufficient short-term benefit has been documented to warrant continued use of this tool.