Four hundred and eighty patients diagnosed with knee osteoarthritis (OA) were randomized to receive electroacupuncture (EA; n=160) or manual acupuncture (MA; n=160) or sham (n=160) for a total of 8 weeks, 3 times/week. The primary outcome of interest was the response rate which consisted of patients improving at least 2 points on the numeric rating scale (NRS) for pain and 6 points on the WOMAC function subscale. Secondary outcomes of interest included the following: NRS pain at weeks 4, 16, and 26; WOMAC function at weeks 4, 16, and 26; WOMAC pain; WOMAC stiffness; quality of life using the 12-item Short Form (SF-12) health survey (i.e., mental health and physical health); incidence of paracetamol; and patient’s global assessment (PGA). WOMAC stiffness, SF-12 , and PGA scores were measured at 4, 8, 16, and 26 weeks follow-up. Also, incidence of adverse events were reported. Results revealed that the response rate, at 8 weeks, was statistically significantly higher in the EA compared to sham (n=0.023) but this was not observed when comparing MA to sham (n=0.051). Furthermore, the response rate at 4, 16, and 26 weeks follow-up was statistically significantly favored in the EA vs sham group but this was observed for MA vs sham only at weeks 16 and 26 (p<0.05). The total incidence of adverse events included 18, 22, and 17 patients in the EA, MA, and sham groups, respectively. NRS scores were statistically significantly improved in the EA vs Sham and MA vs sham groups for all timepoints. WOMAC function and pain were statistically significantly improved in the EA vs sham for all timepoints; yet, for WOMAC function, MA vs sham was only statistically significantly improved at 16 weeks follow-up (p<0.05). For WOMAC pain, a statistical significant improvement in scores was observed for weeks 8,16, and 26 (p<0.05). Moreover, WOMAC stiffness did not statistically significantly differ in MA vs sham groups for all timepoints, but EA vs sham was statistically improved at weeks 8 and 26 (p<0.05). For SF-12 physical health, a statistical significant improvement for EA and MA when compared to sham were observed only at 16 weeks follow-up (p=0.009 for EA vs sham; p=0.026 for MA vs sham). Finally, SF-12 mental health scores statistically significantly favored the EA compared to sham group only at week 26 (p=0.01) whereas MA was favored at both weeks 16 and 26 compared to sham (p=0.024; p=0.001, respectively). Finally, 4, 2 and 10 patients were observed to use paracetamol in the EA, MA, and sham groups, respectively.