We find evidence, by both observation and analysis, that primary crater ejecta play an important role in the small crater (less than a few km) populations on the saturnian satellites, and more broadly, on cratered surfaces throughout the Solar System. We measure crater populations in Cassini images of Enceladus, Rhea, and Mimas, focusing on image data with scales less than 500m/pixel. We use recent updates to crater scaling laws and their constants (Housen, K.R., Holsapple, K.A. [2011]. Icarus 211, 856–875) to estimate the amount of mass ejected in three different velocity ranges: (i) greater than escape velocity, (ii) less than escape velocity and faster than the minimum velocity required to make a secondary crater (vmin), and (iii) velocities less than vmin. Although the vast majority of mass on each satellite is ejected at speeds less than vmin, our calculations demonstrate that the differences in mass available in the other two categories should lead to observable differences in the small crater populations; the predictions are borne out by the measurements we have made to date. In particular, Rhea, Tethys, and Dione have sufficient surface gravities to retain ejecta moving fast enough to make secondary crater populations. The smaller satellites, such as Enceladus but especially Mimas, are expected to have little or no traditional secondary populations because their escape velocities are near the threshold velocity necessary to make a secondary crater. Our work clarifies why the Galilean satellites have extensive secondary crater populations relative to the saturnian satellites. The presence, extent, and sizes of sesquinary craters (craters formed by ejecta that escape into temporary orbits around Saturn before re-impacting the surface, see Dobrovolskis, A.R., Lissauer, J.J. [2004]. Icarus 169, 462–473; Alvarellos, J.L., Zahnle, K.J., Dobrovolskis, A.R., Hamill, P. [2005]. Icarus 178, 104–123; Zahnle, K., Alvarellos, J.L., Dobrovolskis, A.R., Hamill, P. [2008]. Icarus 194, 660–674) is not yet well understood. Finally, our work provides further evidence for a “shallow” size–frequency distribution (slope index of ∼2 for a differential power-law) for comets a few kilometers diameter and smaller.