In the present study the possibility of using nicotine in house dust as an index of environmental tobacco smoke (ETS) exposure was evaluated in an environmental investigation of 23 children with asthma. A standardized procedure for house dust sampling of nicotine with a filter holder connected to a vacuum cleaner, for a defined time and area was developed (F-nicotine). Also, house dust sampling was carried out from the vacuum cleaner bags of the homes (VC-nicotine). There was a larger variation in VC-nicotine (13-655, median 66 microg/g) compared with F-nicotine (15-393 median 156 microg/g). There were statistically significant associations between an inquiry data based ETS exposure index on the one hand, and urinary cotinine concentrations in children (U-cotinine), F-nicotine and VC-nicotine of their homes, on the other. The strong correlation between U-cotinine and F-nicotine (rs = 0.93; P < 0.0001) indicates that the new standardized house dust sampling method should be useful in ETS exposure assessment. However, further validation by a larger sample size with repeated measurements in the same homes is needed.