BACKGROUND: Various types of traction are used in the treatment of low-back pain (LBP), often in conjunction with other treatments. OBJECTIVES: To determine the effectiveness of traction in the management of LBP. SEARCH STRATEGY: We searched The Cochrane Library 2004, Issue 4, MEDLINE, EMBASE, and CINAHL to November 2004, references in relevant reviews, and our personal files. SELECTION CRITERIA: Randomized controlled trials (RCTs) examining any type of traction for the treatment of acute (less than four weeks duration), sub-acute (four to 12 weeks) or chronic (more than 12 weeks) non-specific LBP with or without sciatica. DATA COLLECTION AND ANALYSIS: Study selection, methodological quality assessment and data extraction were done independently by sets of two reviewers. As available studies did not provide sufficient data for statistical pooling, a qualitative analysis was performed. MAIN RESULTS: Twenty-four RCTs, involving 2177 patients (1016 receiving traction) were included in the review. Five trials were considered high quality. There is strong evidence that there is no significant difference in short or long-term outcomes between either continuous or intermittent traction and placebo, sham, or other treatments for patients with a mixed duration of LBP, with or without sciatica. There is moderate evidence that: autotraction is more effective other forms of traction are no more effective than placebo, sham or no treatment for patients with a mixed duration of LBP with sciatica. There is limited evidence that: there is no significant difference in outcomes between a standard physical therapy program with continuous traction and the same program without traction, for patients with a mixed duration of LBP, with or without sciatica autotraction on its own is more effective than a physical therapy program that includes Tru-Trac traction for patients with a mixed duration of LBP with sciatica. There is conflicting evidence regarding the short-term effectiveness of either continuous or intermittent traction compared to placebo, sham or other treatments, in the management of patients who have either chronic LBP or a mixed duration of LBP with sciatica. AUTHORS' CONCLUSIONS: The evidence suggests that traction is probably not effective. Neither continuous nor intermittent traction by itself was more effective in improving pain, disability or work absence than placebo, sham or other treatments for patients with a mixed duration of LBP, with or without sciatica. Although trials studying patients with sciatica had methodological limitations and inconsistent results, there was moderate evidence that autotraction was more effective than mechanical traction for global improvement in this population.