This review looks at the evidence for the effectiveness of spinal cord stimulation in various chronic pain states. Spinal cord stimulation can only be effective when appropriate dorsal column fibres in the spinal cord are preserved and able to be stimulated. Spinal cord stimulation has been shown to have little to offer for patients with some diagnoses. Although 50 to 60% of patients with failed back surgery syndrome obtain significant pain relief with this technique, the strength of the evidence available is insufficient to clearly advocate its use in all patients with this condition. Though limited in quantity and quality, better evidence exists for its use in neuropathic pain, complex regional pain syndrome, angina pectoris and critical limb ischaemia. There is a lack of high quality evidence relating to spinal cord stimulation due to difficulties in conducting randomized controlled trials in this area. Serious methodological problems are encountered in blinding, recruitment and assessment in nearly all published trials of spinal cord stimulation. Suggestions regarding appropriate methodologies for trials which would produce better quality evidence are summarized.