Forty-two subjects suffering from chronic low back pain were matched with the nature of their occupations and then randomly allocated into: (i) an electro-acupuncture group (EA); (ii) an electrical heat acupuncture (EH) group or (iii) a control group. Subjects in the EA group and the EH group received treatment for 20 min on a total of six acupuncture points. Treatment was delivered twice per week for 4 weeks (a total of eight sessions). Back exercise was taught to all subjects, including the control group, as a home programme. A numerical pain rating scale (NPRS), straight leg raise (SLR) and Roland Morris Disability Questionnaire were recorded. There was significant reduction of NPRS within the EA, EH, and control groups across sessions. Significant between-group differences were shown in session 4, session 8 and 1-month follow-up sessions. Post hoc tests showed that the NPRS of the EH group was significantly lower than that of the EA group and the control group by session 4. After session 8, the NPRS of both the EA group and the EH group was significantly lower than that of the control group. Such a difference was maintained at least up to the 1-month follow-up. Only the EA group had significant improvement in the measurement of SLR across sessions. The between-group difference reached significance level in session 8 and at 1-month follow-up. Post hoc tests showed that the EA group had significantly greater gain than the EH group and the control group.