The probability that resuspended roadway dust plays a predominant role in many violations of the ambient particulate standards is now widely accepted. Quantification of this effect is needed before control strategies are developed. In addition, there is increasing concern that strategies should, for reasons of health, also address the respirable fraction of total particulates. This paper describes a twofold approach to the problem of identifying the sources of particulates within the city of Philadelphia and presents selected results from the study. One approach used a sampling program designed to investigate the contribution of traffic-related emissions. Measurements were made using conventional high-volume air samplers at four heights near an intersection and at rooftop level on either side of the main street. The US Environmental Protection Agency dichotomous sampler and the GCA ambient particulate monitor were used to discriminate between respirable and nonrespirable particulates. The second approach used diffusion modeling techniques to calculate the contribution of source categories. Also, a test was conducted to measure the effectiveness of street washing in reducing ambient particulate levels. The results suggest that about one-fourth of the particulates measured at the roof of the Health building were vehicle related, but approximately one-half of the particulates measured at the top of a trailer at a busy intersection were from these sources. Based on a comparison of data from the dichotomous sampler and the high-volume sampler, only about one-fourth of the particulates measured by the high-volume sampler at the trailer were respirable. Ambient particulate levels increased dramatically immediately after street washing./Author/