Episodes with an emotional component preoccupy memory formation and this advantage facilitates their preservation and mitigates the impact of interfering episodes. The present study examined the relation of the emotional dimensions of original and interfering episodes to the memory outcome, using a reconsolidation paradigm. In a between-subjects design, 102 healthy young adults were presented with an emotional or neutral image and learned either an emotional or neutral story, respectively (day 1). On day 2, experimental groups were presented with an image of the opposite emotionality, reactivated the original story, and learned a story of the opposite emotionality. On day 3, experimental and control groups were tested for their memory on target and filler clues of the original story and rated both stories for arousal and valence. Overall, there was evidence of interference on the long-term retention of target clues only for the neutral story (i.e. when the interfering story was emotional), and of filler clues for both types of stories. Moreover, individual target clue retention rates correlated with the arousal ratings for both the original neutral story and the interfering emotional story, while they were not related to arousal ratings for the original emotional story or the interfering neutral one.