The hypothesis of this study was that exposing diesel exhaust (DE*) to the atmosphere transforms its composition and toxicity. Our specific aims were (1) to characterize the gas- and particle-phase products of atmospheric transformations of DE under the influence of daylight, ozone (O3), hydroxyl (OH) radicals, and nitrate (NO3) radicals; and (2) to explore the biologic activity of DE before and after the transformations took place. The study was executed with the aid of the EUPHORE (European Photoreactor) outdoor simulation chamber facility in Valencia, Spain. EUPHORE is one of the largest and best-equipped facilities of its kind in the world, allowing investigation of atmospheric transformation processes under realistic ambient conditions (with dilutions in the range of 1:300). DE was generated on-site using a modern light-duty diesel engine and a dynamometer system equipped with a continuous emission-gas analyzer. The engine (a turbocharged, intercooled model with common-rail direct injection) was obtained from the Ford Motor Company. A first series of experiments was carried out in January 2005 (the winter 2005 campaign), a second in May 2005 (the summer 2005 campaign), and a third in May and June 2006 (the summer 2006 campaign). The diesel fuel that was used closely matched the one currently in use in most of the United States (containing 47 ppm sulfur and 15% aromatic compounds). Our experiments examined the effects on the composition of DE aged in the dark with added NO3 radicals and of DE aged in daylight with added OH radicals both with and without added volatile organic compounds (VOCs). In order to remove excess nitrogen oxides (NO(x)), a NO(x) denuder was devised and used to conduct experiments in realistic low-NO(x) conditions in both summer campaigns. A scanning mobility particle sizer was used to determine the particle size and the number and volume concentrations of particulate matter (PM) in the DE. O3, NO(x), and reactive nitrogen oxides (NO(y)) were monitored using chemiluminescence and Fourier transform infrared instruments. At the end of the exposures, samples of particle-associated and semivolatile organic compounds (SVOCs) were collected from the chamber for chemical analysis using an XAD-coated annular denuder followed by a filter and XAD cartridge. (XAD is an adsorbent polystyrene divinylbenzene resin used in sampling cartridges.) Samples for toxicity testing were collected using Teflon filters followed by two XAD cartridges. The chemical analyses included determination of organic carbon (OC), elemental carbon (EC), carbon fractions, inorganic ions (e.g., sulfate and nitrate), and speciated organic compounds (polycyclic aromatic hydrocarbons [PAHs], nitro-PAHs, polar compounds, alkanes, hopanes, and steranes). The toxicity tests were performed with extracts of PM combined with the SVOCs. The biologic activity of these extracts was evaluated in vivo by instilling them into the tracheas of rodents and measuring pulmonary toxicity, inflammation, and oxidative-stress responses. Results from the chemical analyses indicated that aging DE in the dark with added NO3 radicals and aging DE in daylight with and without additions led to the formation of additional particles and SVOC mass caused by reactions of VOCs, SVOCs, and inorganic gases. The greatest increase in mass occurred with the addition of VOCs as co-reactants. The proportions of the pyrolized OC (POC) fraction increased in the organic mass, which suggested that highly polar and oligomeric compounds had been formed. Results from the toxicity testing were consistent with the hypothesis that the toxicity of the samples had been affected by changes in their composition (caused both by the atmospheric aging and by changes in the initial composition of the DE presumably associated with changes in the engine, which was new at the outset and accrued wear during the study).