Introduction Osteoarthritis (OA) is a common disease with high socioeconomical costs. In Denmark, standard rehabilitation (SR) consists of a combination of patient education and supervised physical exercise involving a standardized neuromuscular training program. As an evidence- based alternative, high-load (>70% 1RM) resistance training (HIRT) has shown positive rehabilitation effects in knee-OA but may not be tolerated in all patients (∼25%) due to knee joint pain. However, low-load resistance training (20-40% 1RM) with concurrent partial blood-flow restriction (BFR) appears to produce effects similar to HIRT yet involving reduced joint pain during and after exercise. The aim is to examine the effect of low-load BFR training compared to SR on pain, thigh muscle mass and muscle function in adults with knee-OA. We hypothesize that 12 weeks of BFR will lead to superior improvements in pain, muscle mass and mechanical muscle function compared to SR. Methods and analysis 90 participants diagnosed with radiographic knee-OA will be randomized to either BFR or SR twice a week for 12 weeks. BFR will consist of two selected lower limb strength exercises performed with an inflated pneumatic occlusion cuff. Intervention procedures in SR consist of a full 8 weeks GLA:D course followed by 4 weeks of team group training. Primary outcome variable is the change in KOOS-Pain subscale from baseline to 12 weeks. Secondary outcome variables are changes in pain sensitivity, functional performance, muscle mass and mechanical muscle function. Intention-to-treat and per-protocol analyses will be conducted. One-way analysis of variance will be performed to evaluate between-group changes. Pre-to-post intervention comparisons will be analyzed using a mixed linear model. Regression analysis will be performed to evaluate potential associations between selected outcome variables.