The health effects of environmental tobacco smoke (ETS) exposure constitute a main public health problem. Lack of presice methods for assessing personal tobacco smoke exposure, makes it difficult to estimate the health effects of such exposure. Measuring hair nicotine concentrations could be an improvement in the assessment of personal tobacco smoke exposure. The objective of the present study was to estimate quantitatively the relation between hair nicotine concentrations in mothers and children and tobacco smoke exposure assessed by questionnaires. Mothers' and children's hair nicotine concentrations in the proximal 2 cm of hair were measured in 94 families with children 12-36 months of age: 25 nonsmoking families, 40 families with one smoking parent, and 29 families with both parents smoking. Questionnaire information on tobacco smoke exposure was collected from the same families. In multivariate linear regression analysis, children's nicotine levels were linearly related to daily number of cigarettes smoked at home by both mothers (0.8 mg/g increase per cigarette, 95% confidence interval [CI] 0.43-1.18), and fathers (1.3, 0.81-1.73). Mothers' nicotine levels were linearly related to both personal smoking (2.7, 1.75-3.55) and fathers' smoking at home (2.1,0.74-3.49). Hair nicotine seems to be a good quantitative measure of exposure to tobacco smoke during the previous months both among active and passive smokers. The non-invasive and simple collection procedure makes the method especially suitable for estimating tobacco smoke exposure in children.