Brittle particle impact attrition was measured over three orders of magnitude of impact velocity by use of both single- and multiple-impact testers. The multiple-impact tester was a resonant cantilever impactor with dynamic amplitude control and small-gap cavities, designed to ensure operation near the transition between the bouncing and resonant impact regimes. For this impactor, a novel technique using a trajectory simulation was developed to elucidate average impact velocities, effective particle restitution behavior, and average losses per impact from a set of nominal attrition rates (loss/time). This allowed direct comparison of data from the single-impact and multiple-impact testers. Results were obtained for three brittle, porous pharmaceutical particles with significantly-different, well-characterized internal structures. Results reveal a relatively unexplored mode of attrition that is distinguished by lack of gross fragmentation at low velocities, with a steep velocity dependence. This regime is expected for any brittle particle for which simple chipping is not seen—such as rounded solid particles and many agglomerates. At higher impact velocity, gross particle fragmentation is observed. The transition between these attrition regimes appears connected to particle structure, such as the size of attrition-resistant primary particles in an agglomerate or the point at which dominant flaws (that lead to fragmentation) are no longer critically active. The ranking of particles according to damage in high-velocity impacts was not predictive of damage in low-velocity impacts, because the particle attrition did not necessarily exhibit the same velocity-dependence in the two regimes. Such differences are critical for predicting performance in operations such as pneumatic conveying or fluidized beds, respectively.