The goal of the “Children's Health Project” is to clarify the association of childhood asthma with elevated concentrations of particulate matter 10 microns and smaller (PM10). To this end, three analytical approaches were applied to metropolitan Phoenix to better understand the spatial distribution and forecasting of PM10: atmospheric modeling, interpolation of measured concentrations, and a statistical predictive system. The emphasis of this paper, which describes the first two approaches, was to develop, evaluate and demonstrate the advantages and limitations of different methods of generating the spatial concentration fields of PM10 that can be linked with the public health data. The MM5/SMOKE/CMAQ deterministic modeling system was employed to simulate the concentration distribution at three nested domains in metropolitan Phoenix and its environs. Design days included low, intermediate and elevated PM10 concentrations. MM5 performance was evaluated with an extensive array of surface meteorological sites; CMAQ's performance, with hourly and daily measured concentrations of PM10 from five monitoring sites with continuous instrumentation in central Phoenix. Both Ordinary Kriging and Inverse Distance Weighting interpolation techniques were used to produce spatial distributions of PM10 from the five permanent continuous monitors, augmented by an additional three months of continuous PM10 monitoring at four temporary sites. The mapped distributions provide the basis for the application of statistical methods to explore the relationship between PM10 concentrations and asthma incidences.