A portable radiance research nephelometer was used to measure the variation in woodsmoke pollution in (a small town of 22,000 people), New South Wales, Australia, on 14 winter nights in 1996. Winter nights are by inversions that trap the air within the valley and reduce winds to very low speeds (averaging 0.15 m s(-1)). varied considerably with location. Mean scattering coefficients (bsp/10 km) for 14 measurement nights ranged from than 1.0 on the undeveloped fringes of the city to 8.7, the latter representing a 14-night average of 200 mu g m(-3) of Pollution was generally highest in the residential areas on either side of the valley, where the smoke was generated, than the low-lying central creeklands. In places, average pollution levels increased 4-fold within 41 m. The between nephelometer and gravimetric pollution measurements ranged 0.95-0.99. The presence of large, sudden repeatable changes in air pollution, and high correlations between nephelometer and gravimetric measurements, that mobile pollution monitoring devices provide a useful and accurate estimate of spatial variability. Estimated for the town as a whole was 1.02 for the 6 months from April to September, 0.25 m October as heater use declines, and 0.12 in normal summer months. For comparison, published 25th, 50th and 75th percentiles of the distribution of coefficients in Sydney were 0. 15, 0.24 and 0.37, respectively. Thus annual exposure to PM2.5 pollution in Armidale woodsmoke is more than double that from all sources in Sydney, a city of 4 million. Overseas estimates of 6% mortality for each additional 10 mu g m(-3) of PM2.5 suggest that wood heaters in Armidale may increase mortality Armidale by about 7%, with estimated cost of about $4270 per woodheater per year. Alternative cheap environmentally friendly methods of keeping houses warm in winter, such as solar heating, should therefore be (C) 2007 Elsevier Ltd. All rights reserved.