The increase in global population witnessed over recent years poses major threats to the quality of the air we breathe. Coupled with population growth in many developed countries, often driven by immigration, there have been substantial increases in the populations of developing countries. At the same time, an increasing number of individuals live in urban environments. In order to assess risk of exposure to hazardous air pollutants (air toxics) in the inner city, where a susceptible population resides, data obtained from high school students in the New York and Los Angeles TEACH (Toxic Exposure Assessment: A Columbia-Harvard Study) studies, were analyzed, with the three main objectives to: 1). Assess, characterize, quantify and compare directly-measured personal air toxics exposures from New York and Los Angeles; 2). Assess, quantify and compare the concentrations and temporal and spatial variability of air toxics measured in the outdoor urban microenvironments of these two cities; 3). Determine if personal air toxics exposures could be modeled using available time-activity information, coupled with measured microenvironmental air pollution inputs. The main findings of this research revealed substantial differences between New York and Los Angeles in the quantity and quality of particle-associated personal air toxics exposures in these two cities. Students across cities exhibited similar levels of personal exposure to particulate matter (PM 2.5) and to high levels of sulfates (greater in Los Angeles than New York; both likely of vehicular traffic origin). Different patterns of exposure to particle-associated air toxics was observed in the two cities and across seasons. In New York, students demonstrated substantially higher exposures to iron, cobalt, and manganese, likely of subway origin, than their counterparts in Los Angeles, who exhibited higher exposures to calcium, aluminum, magnesium (likely of crustal origin). Across seasons, within cities, differences were also detected, with higher levels of air toxics exposures shown in New York in the winter than in the summer for nearly 80% of the analytes, similar to the general pattern (winter vs. fall) in Los Angeles. With respect to outdoor air toxics concentrations, in general terms, crustal sources of air toxics were detected in both cities, though in Los Angeles these levels tended to be higher than in New York, often significantly. Anthropogenic sources were evident in each of the cities to varying degrees. Sulfates were detected at comparable high levels across both cities, though the levels tended to be higher and variability of concentrations of this air pollutant was greater Los Angeles than New York, likely reflecting differences both in patterns of traffic and built environment. Various approaches taken to model spatial and temporal variability of outdoor air toxics concentrations using mixed procedures showed city-specific, spatial and temporal variance patterns of air toxics. Using location and time (day) inputs, in New York, Zn, Pt, and Sn were among the elements with highest spatial variability in the summer, whereas in the winter, Co and La (possibly of subway origin) showed high spatial-temporal variance. In Los Angeles on the other hand, highest spatial to temporal variance ratios were noted for Cs, Ni and K in the fall and Ni, As and Mg in the winter. Each city also revealed different patterns of temporally dominant air toxics, consistent with variable-in-time excursions in air toxics reflecting remote, upwind sources. Using regression modeling that accounted for the distribution of measured personal air toxics, coupled with available time-activity diary data from TEACH and assignment of those activities to specific measured microenvironments, modeling of personal exposures yielded generally strong coefficients of determination, explanatory power and could be cross-validated. Important findings included the role of the indoor environment in predicting personal exposures and the degree to which a small percentage of time spent in the transit environment could affect exposures to trace elements from this source. Although the majority of elements could be predicted in large part by indoor exposures, not simply as a reflection of outdoor air toxics concentrations, the inclusion of other microenvironments, in many cases substantially increased the predictive power of the models generated. The research pursued in this thesis project further details and underscores the risk of air toxics exposures of young residents of the inner city, which, unlike workplace and environmental standards that traditionally may have been based on single exposures, are characterized by exposures to low level complex mixtures of air toxics. In aggregate, these mixtures may have different health consequences than more intense single pollutant exposures. Data generated here may help to inform planning of air quality monitoring approaches in the inner city, as well as provide one template for predictive modeling of human exposures to air toxics in that complex environment, to reduce the need for direct personal measurements to assess exposure risk. This may ultimately contribute to approaches to mitigate air toxics exposures and its consequences for an expanding global population residing in the world's inner cities.