Question: In patients with chronic low-back pain and disc degeneration, how do surgical and nonsurgical treatments compare? Data sources: MEDLINE and references of review articles. Study selection and assessment: Randomized controlled trials (RCTs) that compared surgical with nonsurgical treatment for discogenic back pain. Studies that focused specifically on comparing injections or other percutaneous treatments were not included. Study quality was assessed with use of the checklist for the Consolidated Standards of Reporting Trials. Main outcome measures: Back-specific disability. Main results: 4 RCTs (n = 783; age range, 18 to 65 y) met the inclusion criteria. Surgery in 2 RCTs consisted of posterolateral fusion with use of iliac crest autograft and a specific type of pedicle screw fixation; in 1 RCT, surgeons chose the surgical approach, implant, interbody cages, and bone-graft material; and in 1 RCT, patients were randomized to receive 1 of 3 prespecified surgical techniques: posterolateral fusion with iliac crest autograft and no fixation, posterolateral fusion with iliac crest autograft and pedicle screw fixation, or circumferential fusion consisting of posterolateral fusion and fixation supplemented with interbody fusion with use of autogenous iliac crest bone block inserted anteriorly or posteriorly. One RCT compared surgery with nonoperative care focused on physical therapy, and in 3 RCTs the control treatment was cognitive behavioral therapy addressing fears about back injury. All 4 RCTs measured disability with use of the Oswestry Disability Index (ODI) (higher score = greater disability). Across the 4 RCTs, the percent improvement from baseline ranged from 19% to 37% in the surgery group and from 5.8% to 30% in the nonoperative group. 1 RCT showed a benefit of surgery over unspecified nonoperative treatment with a difference in mean change in the ODI of 8.8 points (percent improvement 19%). The other 3 RCTs did not show a clinically meaningful difference in change in ODI score between surgery and nonoperative treatment (range: in mean change, -3.9 to 3.8%; in improvement, -9.5% to 7.5%). 2 RCTs that reported patient satisfaction showed greater success with surgery. The complication rate with surgery ranged from 9% to 18%; no complications were reported for the nonoperative group in any study. Conclusions: In patients with chronic low-back pain and disc degeneration, surgical treatment may be somewhat better than unstructured nonoperative treatment but has not been shown to be better than cognitive behavioral therapy.