This thesis proposed and tested a novel hypothesis to explain the effect of memory facilitation both at behavioral and brain levels. We propose, based on appraisal theories of emotion, and the autobiographical memory model that goal-relevant information is better remembered than goal-irrelevant information. Depending on the relevance level of these goals, they would be more or less likely to be processed and integrated into long-term memory. From these theoretical frameworks, we refined previous hypotheses explaining emotional memory facilitation. We specified the type of events capable of initiating an arousing response and in turn enhancing memory. We found that neural brain mechanism underlying this memory enhancement showed higher activation in regions related to reward and aversive learning, as well as to emotional processing for goal-relevant contexts as compared to irrelevant contexts. Finally, we were also able to extend the relevance hypothesis to a memory system other than episodic, namely prospective memory.