Objective Depression is a major public health problem. There is some evidence supporting the effectiveness of needle acupuncture in its treatment. Laser stimulation, regarded here as a modality of acupuncture, is non-invasive and therefore free of infection risk; and it is acceptable to patients with needle phobia. The technique is relatively easily learned by community-based general practitioners. It is also amenable to sham control and therefore double-blinding in clinical trials. A double-blind randomised controlled trial was conducted to test the efficacy of low level laser acupuncture in mild to moderate depression. Methods Thirty patients with depression were randomised to receive either active or inactive laser treatment. The laser unit could be switched to one of two settings. One switch position delivered active laser acupuncture and the other was inactive (sham). In the active mode, 0.5J was delivered to each of six to eight individually tailored acupuncture sites per visit. All patients were treated twice weekly for four weeks then weekly for a further four weeks. The patients and the acupuncturist were both blinded to conditions. Outcome was assessed using the Beck Depression Inventory at baseline, weeks four and eight during treatment, and at 4 and 12 weeks following the treatment. Results At the end of the treatment period, Beck Depression Inventory scores fell from baseline by 16.1 points in the intervention group and by 6.8 points in the sham control group (P<0.001). The difference showed only a trend four weeks later, but was again significant after 12 weeks (P=0.007). Laser acupuncture was well tolerated with transient fatigue as the most common adverse effect. Conclusion Laser acupuncture may be worth further investigation as a treatment for mild to moderate depression in primary care.