It is accepted that diet is a major contributor to the obesity epidemic, but environmental ‘obesogenic’ chemicals have also been suggested recently as playing a role, based on in vitro, animal and epidemiological studies. Using two such ‘obesogen’ examples (bisphenol A, certain phthalate esters), we argue that their association with obesity and obesity-related disorders in humans could be circumstantial, and thus non-causal, because a Western style diet increases exposure to these compounds. This possibility needs to be addressed before further (confounded) epidemiological studies on ‘obesogens’ are undertaken.