Objective: To define the association between change in body mass index (BMI) and the incidence and progression of the structural defects of knee osteoarthritis as assessed by radiography. Methods: Radiographic analyses of knees at baseline and at 4 to 5 years’ follow up were obtained from three independent cohort studies: Osteoarthritis Initiative (OAI); Multicenter Osteoarthritis Study (MOST); and Cohort Hip and Cohort Knee (CHECK). Logistic regression analyses using generalised estimating equations, with clustering of both knees within individuals, were used to investigate the association between change in BMI from baseline to 4 to 5 years’ follow up and the incidence and progression of knee osteoarthritis. Results: A total of 9683 knees (from 5774 participants) in an ‘incidence cohort’ and 6075 knees (from 3988 participants) in a ‘progression cohort’ were investigated. Change in BMI was positively associated with both the incidence and progression of the structural defects of knee osteoarthritis. The adjusted odds ratio for incidence was 1.05 (95% confidence interval [CI] 1.02 to 1.09), and for progression was 1.05 (95% CI 1.01 to 1.09). Change in BMI was also positively associated with degeneration (i.e., narrowing) of joint space and with degeneration of the femoral and tibial surfaces (as indicated by osteophytes) on the medial but not on the lateral side of the knee. Conclusion: BMI decrease is independently associated with lower odds of onset and progression of the structural defects of knee osteoarthritis, and could be a component in preventing the onset or worsening of knee osteoarthritis.