in-person instruction on at-home knee exercise. The primary outcome measure was a composite score on the Western Ontario McMaster University Osteoarthritis Index (WOMAC; 100 points), which assessed knee pain, function and stiffness on a per subject basis. The secondary outcome was the Knee Pain Scale (KPS) which assessed knee pain severity and frequency on a per knee basis; both were done at baseline, 5, 9 12, 26 and 52 weeks. Procedure-related opioid medication use, subject satisfaction and adverse events were also assessed. Results: Analysis was by intention to treat. No significant baseline differences existed between the groups in age, gender, pain duration, body mass index or WOMAC scores. 89 subjects (57±8.3 years old, 59 female) with moderate to severe KOA received an average of 4.3±0.7 prolotherapy injection sessions over a 17-week treatment period. All groups reported improved composite WOMAC scores compared to baseline status (p < 0.01) at 52 weeks. However, WOMAC scores for prolotherapy subjects, adjusted for gender, age and body mass index showed significantly greater improvement onWOMAC score at 52 weeks; 15.32±3.52 points for prolotherapy compared to 7.68±3.41 points for saline injection (p < 0.05) and 8.25±3.33 points for exercise (p< 0.05). The improvement by prolotherapy subjects exceeded minimal clinical important difference. KPS scores of prolotherapy subjects showed similar improvement per injected knee compared to baseline status (p < 0.001) and controls (p < 0.05). Prescribed post-procedure opioid medication resulted in rapid diminution of prolotherapy injection pain. Satisfaction with prolotherapy was high and there were no adverse events. Conclusions: Prolotherapy resulted in safe, significant, sustained improvement of pain, function and stiffness scores compared to blinded saline injections and at-home exercises in knee osteoarthritis.