Prenatal exposure to tobacco through maternal smoking during pregnancy increases the risk of many negative health and behavioural outcomes. Across human populations, smoking during pregnancy remains above 10%. To better discover and apply effective interventions, it is important to determine the exposure of each child. Unfortunately, simply asking mothers about their smoking behaviour is not sufficient due to lapses in memory, reluctance to admit smoking or failure to accurately define smoking behaviour. Measurement of cotinine in maternal blood, urine or saliva can improve detection but may miss cases of intermittent smoking because cotinine is cleared quickly during pregnancy. Fortunately, children retain evidence of prenatal tobacco exposure in the DNA methylation of their blood even beyond the age of 17. However, since blood collections are a precious and finite resource many cohort studies have derived lymphoblastoid cell lines (LCLs) from blood samples. LCLs are B cells that have been transformed to spontaneously replicate, thereby providing a constant supply of new biological material. Although the transformation process is known to disrupt DNA methylation marks, a recent study has reported associations between LCL DNA methylation and present tobacco smoking. We would like to test associations of prenatal tobacco exposure in LCL DNA methylation and determine the extent to which the transformation process has modified these associations by comparing them to associations in B cells isolated from matching blood samples. B cell isolation is necessary because each of the many cell types found in blood has a unique DNA methylation profile.