Objective To determine the dose response effect of weight loss on clinical and mechanistic outcomes in overweight and obese adults with knee osteoarthritis (OA). Methods This is a secondary analysis of the diet‐induced weight loss only (D) and diet‐induced weight loss plus exercise (D + E) groups in the Intensive Diet and Exercise for Arthritis randomized controlled clinical trial. The 240 participants were overweight and obese older community‐dwelling adults with pain and radiographic knee OA. Participants were assigned to 1 of 4 groups according to weight loss achieved over an 18‐month period: <5% (<5% group), 5–10% (≥5% group), 10–20% (≥10% group), and >20% (≥20% group). Results There were significant dose responses to weight loss for pain (P = 0.01), function (P = 0.0006), 6‐minute walk distance (P < 0.0001), physical (P = 0.0004) and mental (P = 0.03) health‐related quality of life (HRQoL), knee joint compressive force (P < 0.0001), and interleukin‐6 (P = 0.002). Greater weight loss resulted in superior clinical and mechanstic outcomes, with the highest weight loss group (≥20% group) distinguishing itself on all measures compared with the <5% and ≥5% groups; the ≥20% group had 25% less pain and better function compared with the ≥10% group and significantly (P = 0.006) better physical HRQoL. Conclusion Long‐term weight loss of 10–19.9% of baseline body weight has substantial clinical and mechanistic benefits compared with less weight loss. The value of an additional 10% weight loss includes significantly improved physical HRQoL and a clinically important reduction of pain and improvement in function.