Mechanical shocks induced by the detonation of pyrotechnics ordnances are known to be highly detrimental to the integrity of spacecraft components. As a result, it is desirable to qualify these components as early as possible to prevent failures during launch. Assembly-level shock tests are usually performed on these components. However, these tests are suspected to be the cause of several failures that would not occur during flight. The knowledge acquired in vibration testing suggests that the overtesting is mainly due to the rigidity of the mounting interface between the component and the testing device. Consequently, the test configuration lacks the dynamic absorber effect which should naturally occur in flight configuration. While the Force Limited Vibration (FLV) technique has been developed to reduce the overtesting occurring in vibration testing, no systematic methodology have been proposed to measure and to reduce the shock overtesting to this date.  The objective of this study is to conduct systematic analytical and experimental sensitivity investigation on the overtesting occurring in the assembly-level shock testing. A reconfigurable prototype of an electronic box and a mounting structure have been designed and modeled using the finite element method. The prototype is then fabricated and tested to demonstrate the occurrence of overtesting experimentally. The vibration overtesting and the C 2 coefficients of the semi-empirical method are evaluated using FLV technique. The shock overtesting is evaluated similarly using an acceleration excitation generated for the shaker test. The results show that no significant correlation between the vibration and the shock overtesting exist. Finally, a method to reduce the shock overtesting is proposed in which the shock specification is notched at the main frequencies to limit the response of the structure.