Numerous small- and intermediate-size cratering experiments have been conducted in Piceance Creek Basin oil shale at the Colony and Anvil Points oil shale mines near Rifle, Colorado. The purpose of these experiments was to evaluate scaling as a tool to infer the behavior of large-scale tests from small-scale experiments, to calibrate the hydrodynamic computer codes used to model explosive fragmentation of oil shale, and to investigate the influence of bedding plane orientation, natural joints, fractures, and the grade of oil shale on rock fragmentation. The small tests were made using PETN and RDX explosive with charge sizes of a few grams. The intermediate-sized tests used ANFO or TNT explosives with charge sizes of 5 to 100 kg. Crater dimensions were measured on all experiments. Crater volumes were calculated from screened rubble volumes on the intermediate-scale experiments and measured directly on the small-scale experiments. Fragment size distributions were measured on most of the intermediate-sized tests and on several of the small-scale experiments. The analyses of these cratering data show: (1) small-scale cratering tests can be used to qualitatively predict the kinds of geologic interactions that will influence a larger-scale experiment; (2) the site specific geology plays a dominant role in the formation of the crater; (3) small flaws and fractures influence crater development and particle size distributions in small-scale craters in the same manner that joint and fracture systems influence intermediate-scale experiments; (4) complex site geology causes increases in the critical and optimum depths of burial and changes the symmetry of the crater; and (5) small- and intermediate-scale cratering experiments can be used to calibrate hydrodynamic computer codes if great care is used to identify the effect of site specific geology. 15 figures, 6 tables.