Abstract Background Knee Osteoarthritis (KOA) is a very common condition with prevalence rising with age. It is a major contributor to global disability and has a large socioeconomic burden worldwide. Conservative therapies have marginal effectiveness, and surgery is reserved for severe symptomatic KOA. Dextrose Prolotherapy (DPT) is an evidence-based injection-based therapy for chronic musculoskeletal conditions including KOA. The standard “whole joint” injection method includes intra-articular injection and multiple extra-articular injections at soft tissue bony attachments. The procedure is painful and requires intensive procedural training often unavailable in conventional medical education, which potentially limits access. Intra-articular injection offers the possibility of a less painful, more accessible treatment. The aim of this project is to assess the clinical efficacy of intra-articular injection of DPT versus normal saline (NS) for KOA. Method Seventy-six participants with KOA will be recruited from the community. We will conduct a single center, parallel group, superiority randomized controlled trial comparing DPT and NS injections, with blinding of physician, participants, outcome assessors and statisticians. Each group will receive injections at week 0, 4, 8 and 16. The primary outcome will be the Western Ontario McMaster University Osteoarthritis Index pain scale (WOMAC), and secondary outcomes include WOMAC composite score, the WOMAC function and stiffness subscale, the Visual Analogue Score of pain, objective physical function tests (the 30 s chair stand, 40- m fast paced walk test, the Timed up and go test) and the EuroQol-5D (EQ-5D). All outcomes will be evaluated at baseline, and 16, 26 and 52 weeks. All analyses will be conducted on an intention-to-treat basis using linear mixed regression models. Discussion This paper presents the rationale, design, method and operational aspects of the trial. The findings will determine whether IA DPT, an inexpensive and simple injection, is a safe and effective non-surgical option for KOA. The results can be translated directly to clinical practice, with potentially substantial impact to patient care. Trial registration The trial (ChiCTR-IPC-15006617) is registered under Chinese Clinical Trials Registry on 17th June 2015.