Numerous research in both human and non-human domains have explored the topic of how expectancy violation affects memory by triggering memory reconsolidation. These violations have shown enhanced efficacy in exposure therapies. However, reconsolidation intervention success is highly dependent on the destabilization of memory and hence cannot be guaranteed. Determining the prerequisites for instability is one of the important problems in the area. In this study, our goal was to demonstrate how prediction error and memory strength (Reinforcement level) affect destabilization of memory. We generate prediction error using UCS reactivation. Sixty participants were tested on a three-day fear conditioning paradigm: acquisition (50% and 100%) on Day 1; UCS Reactivation or control on Day 2; extinction learning and reinstatement on Day 3 where data was recorded using SCR and subjective measures. We found that UCS reactivation could generate greater expectancy violation even for stronger memories. The failure of fear return can be explained by reduced valency of UCS as found in the valence ratings and increased discrepancy in expectations. Hence, the study indicates that memory reconsolidation depends on the strength of fear memory as well as the prediction error generated. The study further helps in increasing therapeutic efficacy of anxiety disorders and Post-Traumatic Stress Disorder.