The deviations of trajectory and attitude angle for internal store separation are evaluated by two wind tunnel test methods. One is the Freedrop Test (FDT), which is known as unsteady and time-dependent method of scaled model. The other is the Captive Trajectory System (CTS) test, which is usually regarded as a quasi-steady and time-averaged test technology. The result shows that there is a streamwise adverse pressure gradient on the cavity resulting in a nose-up pitching moment coefficient (>0) acting on the store model. When the initial pitch angular velocity is 0, the store exits the shear-layer with a nose-up pitch attitude causing the store to climb back towards and collide with the aircraft. However, the store passes through the shear-layer into the freestream with a nose-down pitch, which causes a successful separation event when the initial pitch angular velocity < 0. The pitch angle obtained by unsteady test method is different from that by quasi-steady test method. The time-dependent test includes the aerodynamic force induced by pitch angular velocity whereas the time-averaged method (CTS) cannot reflect the effect of unsteady aerodynamic force. The deviation of vertical displacement is not obvious for FDT and CTS test since the store has an initial vertical velocity, which is dominant for the vertical displacement. This means that the highly unsteady flow can create unpredictability in aerodynamic pitching moment of the store, which can lead to the deviation of pitch angle for internal store separation.