This dissertation examined the contributions of mood valence and arousal to false memory (FM) for neutral DRM word lists. There has been a lively debate in the literature regarding whether valence or arousal is the primary cause of mood-dependent FM. Because much past research has not effectively disentangled valence from arousal, the present studies used a new mood manipulation that allowed valence and arousal to be controlled and their associated retrieval processes to be measured. In Experiment 1, low-arousal negative moods reduced false memory compared to positive and neutral moods by increasing verbatim memory traces, confirming the prediction that valence can influence false memory independent of arousal. In Experiment 2, low-arousal negative moods reduced false memory for both list lengths, but the processes associated with false memory for short lists were not the same as the processes associated with false memory for long lists. High arousal negative moods in Experiment 3 now increased false memory compared to neutral moods—but still reduced false memory compared to positive moods—which suggested that high arousal increased false memory compared to low arousal. .Experiment 4 provided a direct test of arousal’s effect on false memory and confirmed that higher levels of arousal increase false memory via strong gist traces, regardless of valence. The results were discussed in support of fuzzy-trace theory—but not affect-as-information theory—and practical applications to the law were noted.