A double blind randomised controlled clinical trial on the effect of transcutaneous spinal electroanalgesia (TSE) on low back pain was carried out in 58 patients attending a Pain Management Unit. Four TSE instruments, two active and two sham, were used and each patient was assigned randomly to one of these. Low back pain was rated by each patient using a visual analogue scale (VAS) immediately before and immediately after a single 20 min treatment of TSE and also daily for the week prior to, and the week following, the treatment. No significant difference in mean pain score was detected between the active and sham treated groups immediately after treatment or during the subsequent week. The Hospital, Anxiety and Depression scale (HAD) and the General Health Questionnaire (GHQ) were completed by each patient and there was a positive correlation between the scores achieved on these scales and the mean pain scores in both the active and sham treated groups. A post-trial problem was the discovery that the specification of the two active TSE machines differed from the manufacturer's specification. Thus, the output frequencies were either more (+10%) or less (-17%) while the maximum output voltages were both less (-40% and -20%), respectively. However, additional statistical analysis revealed no significant differences between the results obtained with the two active machines.