The UK Building Research Establishment has carried out measurements of levels of nitrogen dioxide (NO2) in 12 homes in the South of England. Two types of detection device were used: the Scintrex LMA-3 continuous NO2 analyser and the Palmes passive diffusion tube. NO2 concentrations were recorded using both devices in the kitchen, living room and a bedroom of each home for a period of a week. Gas cooking was observed to be the most important source of peak concentrations of NO2 in the home. Six homes, out of the ten that used gas for cooking, had levels of NO2 which exceeded the WHO 1-hr guideline value of 210ppb. In at least two of the homes, personal exposure is expected to have also approached or exceeded this guideline value. There was a fairly good correlation of measured average levels of NO2 between the two detection devices. There was also a fairly good correlation between the 7-day average levels recorded by the diffusion tubes and the maximum 1-hour averages recorded by the continuous detectors. This offers the possibility of gaining information on short-term exposure levels from long-term diffusion tube data. An investigation of the effect of a cooker hood on NO2 levels in a kitchen suggested that it could prove to be an effective means of removing NO2 generated during cooking, especially if there are no other competing means of ventilation such as open doors.