The Capstone Depleted Uranium (DU) Aerosol Study, which generated and characterized aerosols containing DU from perforation of armored vehicles with large-caliber DU penetrators, incorporated a sampling protocol to evaluate particle size distributions. Aerosol particle size distribution is an important parameter that influences aerosol transport and deposition processes as well as the dosimetry of the inhaled particles. These aerosols were collected on cascade impactor substrates using a pre-established time sequence following the firing event to analyze the uranium concentration and particle size of the aerosols as a function of time. The impactor substrates were analyzed using proportional counting, and the derived uranium content of each served as input to the evaluation of particle size distributions. Activity median aerodynamic diameters (AMADs) of the particle size distributions were evaluated using unimodal and bimodal models. The particle size data from the impactor measurements were quite variable. Most size distributions measured in the test based on activity had bimodal size distributions with a small particle size mode in the range of between 0.2 and 1.2 microm and a large size mode between 2 and 15 microm. In general, the evolution of particle size over time showed an overall decrease of average particle size from AMADs of 5 to 10 microm shortly after perforation to around 1 microm at the end of the 2-h sampling period. The AMADs generally decreased over time because of settling. Additionally, the median diameter of the larger size mode decreased with time. These results were used to estimate the dosimetry of inhaled DU particles.