OBJECTIVE The purpose of this study was to determine the extent to which various levels of restrictions on smoking in the home may be associated with children's exposure to environmental tobacco smoke (ETS). METHODS The methodology consisted of a cross-sectional survey involving 249 children with asthma aged 1 to 11 attending hospital outpatient clinics, with at least one parent who smoked, linked to the child's urinary cotinine to creatinine ratios (CCR). RESULTS After adjustment for child's age, mother's smoking status, and total parental daily cigarette consumption, a total ban was associated with significantly lower urinary CCR levels (7.6 nmol/mmol) than bans with exceptions or limited smoking in the home. Where exceptions to bans were made (14.9 nmol/mmol), children's urinary CCR levels were no different from homes in which smoking was allowed in rooms the child rarely frequented (14.1 nmol/mmol). These two intermediate levels of restriction were in turn associated with significantly lower CCR levels than unrestricted smoking in the home (26.0 nmol/mmol). CONCLUSIONS Making exceptions to bans on smoking at home measurably undermines the protective effect of a ban. However, making some exceptions to a ban and limiting smoking to rooms where the child rarely goes may result in reduced exposure to ETS, compared with unrestricted smoking.