Sixty patients aged 60 or over with back pain for at least 6 months were recruited from General Practitioner referrals and randomized to 4 weeks of treatment with acupuncture or transcutaneous electrical nerve stimulation (TENS). All treatments were administered by the same physiotherapist and both groups had the same contact with him. The following were measured at baseline, completion and at a 3-month follow-up by an independent observer blinded to treatment received: (1) pain severity on visual analogue scale (VAS); (2) pain subscale of Nottingham Health Profile (NHP); (3) number of analgesic tablets consumed in previous week; (4) spinal flexion from C7 to S1. Thirty-two patients were randomized to acupuncture and 28 to TENS; only three withdrew (two from acupuncture, one from TENS). Significant improvements were shown on VAS (P < 0.001), NHP (P < 0.001) and tablet count (P < 0.05) between baseline and completion in both groups, these improvements remaining significant comparing baseline with follow-up with a further non-significant improvement in VAS and NHP in the acupuncture group. The acupuncture but not the TENS patients showed a small but statistically significant improvement (P < 0.05) in mean spinal flexion between baseline and completion which was not maintained at follow-up. Thus in these elderly patients with chronic back pain both acupuncture and TENS had demonstrable benefits which outlasted the treatment period. Acupuncture may improve spinal flexion. This trial cannot exclude the possibility that both treatments are 'placebos'.