During the Capstone Depleted Uranium (DU) Aerosol Study, aerosols containing DU were produced inside unventilated armored vehicles (i.e., Abrams tanks and Bradley Fighting Vehicles) by perforation with large-caliber DU penetrators. These aerosols were collected and characterized, and the data were subsequently used to assess human health risks to personnel exposed to DU aerosols. The DU content of each aerosol sample was first quantified by radioanalytical methods, and selected samples, primarily those from the cyclone separator grit chambers, were analyzed radiochemically. Deposition occurred inside the vehicles as particles settled on interior surfaces. Settling rates of uranium from the aerosols were evaluated using filter cassette samples that collected aerosol as total mass over eight sequential time intervals. A moving filter was used to collect aerosol samples over time, particularly within the first minute after a shot. The results demonstrate that the peak uranium concentration in the aerosol occurred in the first 10 s after perforation, and the concentration decreased in the Abrams tank shots to about 50% within 1 min and to less than 2% after 30 min. The initial and maximum uranium concentrations were lower in the Bradley vehicle than those observed in the Abrams tank, and the concentration levels decreased more slowly. Uranium mass concentrations in the aerosols as a function of particle size were evaluated using samples collected in a cyclone sampler, which collected aerosol continuously for 2 h after perforation. The percentages of uranium mass in the cyclone separator stages ranged from 38 to 72% for the Abrams tank with conventional armor. In most cases, it varied with particle size, typically with less uranium associated with the smaller particle sizes. Neither the Abrams tank with DU armor nor the Bradley vehicle results were specifically correlated with particle size and can best be represented by their average uranium mass concentrations of 65 and 24%, respectively.