Velocities of ejecta from seven impacts of aluminum projectiles into coarse‐grained sand have been measured with a laser‐based apparatus that produces stroboscopic photographs of individual grains in ballistic flight. Speeds and angles of the majority of the ejecta can then be measured very precisely. There appears to be little effect of impact velocity on the functional relationship between the scaled, radial launch position and either the speed or angle of ejection; the seven experiments covered a range of impact velocities from 0.8 to 1.9 kms^—1. The measured ejection speeds follow power‐law distributions, as predicted by dimensional analysis, but the angle of ejection is not constant throughout a given event as predicted. Indeed, the angle of ejection declines gradually with increasing radial distance from the impact point, but there are indications that the angle increases again as the position of the final crater's rim is approached. The exponents determined from scaled crater dimensions and ejection‐speed distributions are substantially different. Although this might imply that assumptions used in the dimensional analysis are not valid, it is also possible that the coarse sand, whose component grains were comparable in dimension to the diameter of the impactors, instead presented a target that was more of an inhomogeneous aggregate of large fragments than a uniform, continuous medium.