Abstract Shoulder pain is the third most common musculoskeletal problem and accounts for 5% of general practitioner consultations. Although many treatments are described, there is no consensus on optimal treatment and up to 40% of patients still have pain 12 months after initially seeking help for pain. Previously, the effect of transcutaneous pulsed radiofrequency treatment (TCPRFT) was evaluated in a retrospective audit that showed good pain relief for a mean 395 days and justified this randomized sham controlled trial. In this study, 51 patients entered into a randomized double-blinded, placebo controlled study of TCPRFT. Patients were assessed at 4 and 12 weeks by a blinded observer and compared with baseline. We observed sustained reductions in pain at night, pain with activity, and functional improvement at 4 and 12 weeks with active but not sham TCPRFT. The 25 subjects who received active treatment showed statistically significant reductions of 24/100 in pain at night and 20/100 of pain with activity at 4 weeks and 18/100 and 19/100, respectively, at 12 weeks from baseline. Statistically significant lower Brief Pain Inventory pain and function scores (4 and 12 weeks), improved pain self-efficacy (4 weeks), Oxford Shoulder scores (12 weeks), and internal rotation (12 weeks) were seen. Pain at both rest and shoulder elevation were not improved by active treatment. No complications were seen. This study of a simple, low risk, outpatient treatment confirms the findings of our earlier study of TCPRFT for knee pain and shoulder pain audit that transcutaneous pulsed radiofrequency treatment may help some people with painful shoulders.