The gravity flow of crushed Colorado oil shale from a 18 cm dia. glass column through flat circular and converging conical orifices was studied using high-speed cinematography. Factors tested were: (1) diameter of flat-entry and conical orifices, (2) three sizes of shale in one broad and two narrow distributions, (3) upward air velocity, (4) viscosity and surface tension of wetting liquids (one simulating hot shale oil), and (5) presence of up to three crossing tubes in the shale bed. Arching sometimes caused stoppage of flow with orifice-to-particle size ratios (D/W) as high as 7.25. No interparticle motions were visible at levels above about 1.5 column diameters from the orifice. Except when there was air flowing upward, discharge rates were not consistently different when the bed was full than at other levels while it emptied, but local mean velocities measured over 10 cm depth sections ranged widely, about 40% of the mean value. For the runs made with dry shales through flat orifices, the rate of discharge was proportional to the square of the effective opening, D-2W, rather than the 2.5 power reported by other workers for more uniform, smoother, rounder materials. In a flat-bottom vessel, crushed shale forms its own final entrance cone around the orifice, with an included angle of 90°. When conical throats with an included angle of 40° were inserted, flow rates were 50–100% higher. Shale 6 mm and larger that was wet with simulated shale oil flowed as well as dry shale. The presence of only 0.7% of very fine material in dry shale appears to have a lubricating effect, causing the shale to pack more tightly and flow faster.