Systematic review. To systematically assess benefits and harms of nonsurgical interventional therapies for low back and radicular pain. Although use of certain interventional therapies is common or increasing, there is also uncertainty or controversy about their efficacy. Electronic database searches on Ovid MEDLINE and the Cochrane databases were conducted through July 2008 to identify randomized controlled trials and systematic reviews of local injections, botulinum toxin injection, prolotherapy, epidural steroid injection, facet joint injection, therapeutic medial branch block, sacroiliac joint injection, intradiscal steroid injection, chemonucleolysis, radiofrequency denervation, intradiscal electrothermal therapy, percutaneous intradiscal radiofrequency thermocoagulation, Coblation nucleoplasty, and spinal cord stimulation. All relevant studies were methodologically assessed by 2 independent reviewers using criteria developed by the Cochrane Back Review Group (for trials) and by Oxman (for systematic reviews). A qualitative synthesis of results was performed using methods adapted from the US Preventive Services Task Force. For sciatica or prolapsed lumbar disc with radiculopathy, we found good evidence that chemonucleolysis is moderately superior to placebo injection but inferior to surgery, and fair evidence that epidural steroid injection is moderately effective for short-term (but not long-term) symptom relief. We found fair evidence that spinal cord stimulation is moderately effective for failed back surgery syndrome with persistent radiculopathy, though device-related complications are common. We found good or fair evidence that prolotherapy, facet joint injection, intradiscal steroid injection, and percutaneous intradiscal radiofrequency thermocoagulation are not effective. Insufficient evidence exists to reliably evaluate other interventional therapies. Few nonsurgical interventional therapies for low back pain have been shown to be effective in randomized, placebo-controlled trials.