Participants were presented with targets on two occasions. Each time, participants were asked to generate a past episode associated with the targets and judge the pleasantness of that episode during an orienting task, followed by an unexpected free recall test. For participants with high emotional intelligence (EI), the spacing effect (i.e., the advantage of spaced over massed presentation) was observed for targets associated with both pleasant and unpleasant episodes. For participants with low EI, that effect was found for only targets with unpleasant episodes. These results suggest that pleasant episodes elicited weaker emotional encoding than unpleasant episodes did. Furthermore, participants with high EI could activate emotional encoding elicited by pleasant episodes even if the emotion was not strongly experienced.