It has been recommended to use biomarkers in an antenatal setting when focusing on smoking and pregnancy research (Delpisheh et al, 2007). This is primarily due to biomarkers being considered more objective than using questionnaires to measure levels of smoking in pregnant women. However, when a study was designed using a biomarker to compare the levels of smoking reported by pregnant women at the booking appointment, numerous confounding issues became apparent. This article aims to introduce midwives to common biomarkers frequently used in smoking and pregnancy research and to discuss confounding issues that can affect a biomarker's performance using cotinine as a prime example. The confounding factors are discussed under the themes of population, metabolism, specimen, tobacco and behaviour.