Twenty-four adults with symptomatic knee osteoarthritis were randomized to knee joint distraction (KJD) or knee arthroplasty in a multicentre, pragmatic, phase III non-inferiority randomized controlled trial. The primary outcome was the Knee injury and Osteoarthritis Outcomes Score (KOOS) pain subscale at 12 months. Secondary outcomes included other KOOS subdomains, pain, Oxford Knee Score, range of motion, timed up-and-go, complications, and health economic outcomes. Outcomes were planned up to 24 months, but the trial was terminated early due to cessation of elective surgery during the COVID-19 pandemic, limiting follow-up and power. Overall, results showed that both groups experienced improvements in pain and function from baseline to 12 months. However, the small sample size precluded meaningful between-group comparisons or confirmation of non-inferiority. KJD was associated with pin-site complications, while arthroplasty showed larger absolute improvements in several patient-reported outcomes. These findings suggest that while KJD may improve symptoms, its comparative clinical and cost-effectiveness relative to arthroplasty remains uncertain.