During initial shock testing of an electro-optical device weighing approximately 100 lb, a number of catastrophic failures occurred under the shock test conditions specified for simulating the effects of firing the main armament of a tank. In addition, optical boresight shifts and permanent deformations of the cast-aluminum equipment housing were observed. At the same time, three identical devices which had been exposed in the field to a significant number of gunfirings in the tank exhibited no failures of either type. This paper discusses some of the problems which were examined during an investigation of the gunfire shock environment and the means by which its effects, or rather, absence of effects, could be simulated in the laboratory. The intent of the paper is not to break new ground but rather to discuss the several interacting parameters which enter into the simulation process and indicate the implications of the choice of values for these parameters.