A systematic literature synthesis was performed to analyze the long-term risks and benefits of spinal cord stimulation for patients with failed back surgery syndrome. Relevant articles were identified through a MEDLINE search (January 1966-June 1994), bibliography reviews, searches of personal files, and literature supplied by a stimulator manufacturer. Two investigators independently reviewed each article to determine whether it met the following study inclusion criteria: 1) original data on return to work, pain, medication use, reoperations, functional disability, or stimulator use after permanent implantation of spinal cord stimulators in patients with chronic low back or leg pain despite previous back surgery; and 2) follow-up > or = 30 days for all patients. Articles were excluded if data from patients with other diagnoses were mixed with (and could not be separated from) data from patients with chronic low back or leg pain, or if their data were redundant with those reported in an included article. Articles written in languages other than English or French were excluded. Thirty-nine studies, all case studies, were analyzed. At follow-up (mean, 16 mo; range, 1-45 mo), an average of 59% of patients had > or = 50% pain relief (range, 15-100% of patients). Complications occurred in 42% of patients but were generally minor. It seems that approximately 50 to 60% of patients with failed back surgery syndrome report > 50% pain relief with the use of spinal cord stimulation at follow-up; the lack of randomized trials precludes conclusions concerning the effectiveness of spinal cord stimulation relative to other treatments, placebo, or no treatment.