We found an inverse association between cigarette smoking and risk of knee OA, irrespective of study design. This association was more apparent in males. However, we have not demonstrated a causal relationship between smoking and OA, and further investigations are needed.