There is a certain controversy over whether there is a relationship between passive smoking and bronchogenic cancer. The first epidemiology studies has found a small increase in the relative risk. Experimental work in animals have demonstrated that the smoke produced by a burning cigarette contains more cancerogenic agents than the smoke inhaled by a smoker. In man where only epidemiology studies are possible, the problem lies in quantifying exposure to passive smoking. Questionnaires or assay of markers such as CO and HbCO, thiocyanates, nicotine and its metabolite cotinine and certain adduits have been used. There are many possibilities for bias in such studies including classification of smokers among non-smokers, confounding factors (age, profession, eating habits), subjects' memory, and data collection errors. Meta-analyses have attempted to overcome the lack of power in many studies. Some of these meta-analyses have used correction factors to take into account publication bias. Meta-analyses have led to the conclusion that the risk for women exposed to their husband's tobacco smoke is somewhat increased compared with a non-smoking non-exposed population. The relative risk is about 2-fold higher and has been observed in Greece. The existence of a dose-intensity relationship has not been confirmed. Likewise, it is impossible to conclude concerning the role of passive smoking in children due to the small amount of data available.