Cigarette smoking in pregnancy is strongly associated with alterations in offspring DNA methylation profile. This motivates analysis to investigate whether DNA methylation modified by tobacco smoke exposure in utero plays a causal role in mediating its adverse effect on offspring health outcomes, including compromised lung function. Traditional methods for mediation analysis, including the extent to which the association between the exposure and outcome is attenuated with adjustment for the mediator, have several limitations. Alternative approaches have been developed to address some of these limitations, including novel approaches based on counterfactual theory and Mendelian randomization, and an appraisal of these methods in an applied context is warranted.