Retrieval-induced forgetting refers to the finding that repeated retrieving a subset of previously studied items can cause forgetting of related non-retrieved items. Results from prior work suggest that emotions can influence how information is processed. There is evidence that negative emotions result in predominantly item-specific processing and positive emotions result in predominantly relational processing. Retrieval-induced forgetting which presupposes relational processing has been found to be eliminated if individuals process events in an item-specific way. Thus, positive and negative emotions may have opposing effects on retrieval-induced forgetting. Researches extending the procedure to autobiographical memory found forgetting may occur in both neutral and emotional autobiographical memories. However, research concerning the effects of mood on retrieval-induced forgetting has been argued that the experience of negative moods during retrieval can eliminate the later forgetting. The current study is to explore the influence of emotion on retrieval-induced forgetting from two dimensions. One is whether retrieval-induced forgetting would be different for emotional and neutral contents; the other is whether the affective state experienced during retrieval can influence forgetting. To examine whether the emotionality of material would affect retrieval-induced forgetting, experiment 1 selected sixty positive, negative and neutral words of three categories from the Chinese Affective Words System. Twenty-one students studied these emotional words and then practiced retrieving some of words for some of the categories. After a 1-minute distracter task, they tried to recall all these words. Recall performance was then compared for three types of items. To examine whether mood can modulate retrieval-induced forgetting, experiment 2 constructed three word lists, each of them including three semantic categories. Every category included six neural exemplars. Twenty-one students studied these materials and then were asked to repeatedly retrieve a subset of the material. Immediately before retrieval, positive, negative, or neutral moods were induced. Subjects were successively shown five pictures of the same valence which were selected from the Chinese Affective System, and were instructed to let the pictures influence their emotional state. In the final test, subjects were provided all category names to recall all words. Recall performance for words that received retrieval practice were reliably higher on final test than baseline level, whereas recall performance for words that were not practiced but belonging to the same categories as the practiced words were reliably lower than baseline level. In experiment 1, retrieval practice caused the same amount of forgetting for both neutral and emotional words, indicating that the emotionality of material does not affect retrieval-induced forgetting. This result was consistent with prior research. In experiment 2, when subjects were in positive and negative moods, repeated retrieval did not cause forgetting of the nonretrieved material, whereas when subjects were in neutral moods, they showed reliable retrieval-induced forgetting. The findings suggest that the affective state experience during retrieval can modulate retrieval-induced forgetting, that is, retrieval cannot induce forgetting in positive mood, whereas negative mood can eliminate such forgetting. The two experiments tested the effect of emotion on retrieval-induced forgetting from two related but separate dimensions. The results suggest that the emotionality of material does not affect retrieval-induced forgetting, but the affective state experience during retrieval can modulate such forgetting. These findings illustrate that the influence of emotion on retrieval-induced forgetting is predominantly the effect of affective state. Moreover, the present results which imply mood can affect retrieval-induced forgetting may provide information on the interplay between mood and episodic forgetting.