Four treatments for soft-tissue rheumatism--sham ionisation, placebo ionisation, and pharmacological ionisation with pirprofen (two-dose levels)--were assessed in a randomized double-blind, between-patient controlled trial in 73 outpatients affected by scapulo-humeral periarthritis or elbow epicondylitis. Treatment lasted two weeks (5 sessions a week). Progress was measured by patient's assessment on pain at rest and on movement and by physician's assessment on functional impairment. At two weeks each treatment was associated with a significant degree of improvement; however, pharmacological ionisation produced a significantly higher improvement in symptoms. No differences were detected between sham ionisation and placebo ionisation. These results suggest that the ionisation procedure displays per se a moderate therapeutic effect which seems to be due more to a simple placebo effect than to the biological effect of electricity.