QUESTION: Which interventions for non-specific neck pain are effective in reducing pain or disability? DESIGN: Systematic review with meta-analysis of randomised controlled trials. PARTICIPANTS: Adults with non-specific neck pain. INTERVENTION: All interventions for neck pain that were evaluated in trials with a placebo, minimal- or no-intervention control. OUTCOME MEASURES: Pain and disability outcomes (0-100 scale) at the conclusion of a course of treatment (short term), and in the medium (3 to 9 months) and long (> 9 months) term. RESULTS: 33 trials were identified. The interventions with significant short-term effects on pain were manipulation (MD -22, 95% CI -32 to -11), multimodal intervention (MD -21, 95% CI -34 to -7), specific exercise (MD -12, 95% CI -22 to -2), combination orphenadrine/paracetamol (MD -17, 95% CI -32 to -2), and manual therapy (MD -12, 95% CI -16 to -7). There was a significant short-term effect on disability for acupuncture (MD -8, 95% CI -13 to -2) and manual therapy (MD -6, 95% CI -11 to -2). Treatment with laser therapy resulted in better pain outcomes at medium-term follow-up but not at short-term follow-up. No other intervention demonstrated medium- or long-term effects. CONCLUSION: Some conservative interventions for neck pain are effective in the short term. Few interventions that have been investigated have shown longer term effects that are better than placebo or minimal intervention.