This work developed a passive sampling method to measure ambient PM10-2.5. The coefficient of variation (CV) of PM10-2.5 measured with collocated passive samplers was 20.1% in laboratory tests and 11.6% in field tests. PM10-2.5 measured passively deviated from that measured with a filter-based dichotomous sampler by 29% in field tests. The strong correlation between PM10-2.5 derived passively with the dichotomous sampler (r = 0.97) suggests that the bias observed in field tests was systematic and may be empirically corrected. The limit of detection of the method was determined to be 2.3 mu g m(-3) for a 5-day sample. A bootstrap analysis suggests that at least 300 particles should be imaged to stabilize the analytical variability of the method. Repeated analysis suggests that the CV for the analytical method is 5%. This method offers an inexpensive means to assess PM10-2.5 that may be incorporated into compliance networks or epidemiological studies. (C) 2007 Elsevier Ltd. All rights reserved.