Considerable evidence has been reported for wall-rock velocities during rockbursts in deep gold mines that are substantially greater than ground velocities associated with the primary seismic events. Whereas varied evidence suggests that slip across a fault at the source of an event generates nearby particle velocities of, at most, several m/s, numerous observations, in nearby damaged tunnels, for instance, imply wall-rock velocities of the order of 10 m/s and greater. The common observation of slab buckling or breakouts in the sidewalls of damaged excavations suggests that slab flexure may be the mechanism for causing high rock ejection velocities. Following its formation, a sidewall slab buckles, causing the flexure to increase until the stress generated by flexure reaches the limit 5 that can be supported by the sidewall rock. I assume here that S is the uniaxial compressive strength. Once the flexural stress exceeds S, presumably due to the additional load imposed by a nearby seismic event, the slab fractures and unflexes violently. The peak wall-rock velocity v thereby generated is given by v=(3 + 1-??2/2)1 2 S/?????E for rock of density ??, Young's modulus E, and Poisson's ratio ??. Typical values of these rock properties for the deep gold mines of South Africa yield v= 26 m/s and for especially strong quartzites encountered in these same mines, v> 50m/s. Even though this slab buckling process leads to remarkably high ejection velocities and violent damage in excavations, the energy released during this failure is only a tiny fraction of that released in the primary seismic event, typically of magnitude 2 or greater.