Previous studies of embodiment effects have suggested that these effects are directly induced by sensory-motor experience. However, this explanation overemphasizes the influence of body states and ignores modulation from any kind of cognitive factors, which limits the explanation of the boundary conditions and mediators of embodiment effects. One of the most well-known embodiment effects is caused by the sensation of weight. Many studies have found that the experience of weight can subconsciously impact complex cognitive activities. This study investigated the influence of weight experience on the processing of moral concept, that is the “weight-morality” embodiment effect, and especially focused on the modulation effect of the metaphorical association. A joint-classification response task was used to test the embodiment effect. The classification task included two blocks, and 80 trials in each block tested one joint-classification condition (i.e. "heavy-moral, light-immoral" vs. "light-moral, heavy-immoral" condition). In each trial, followed the fixation was a Chinese two-character word with moral or immoral valence, presented up to 3 seconds in the center of the monitor. Two identical computer mouse devices with different weights (70g vs. 200g) were used for participants to experience weight sensation and to classify the words, with the devices placed in their left and right hand respectively. Participants were required to accurately classify the moral words presented in the monitor as quickly as possible, clicking the light or the heavy mouse according to the instruction. There were breaks in each condition for the participants to exchange the mouse devices between their two hands and rest their arms. Event-related potential (ERP) data were collected and we focused on the late positive component (LPC), which has been found that the motivationally salient congruent trials elicit significantly larger LPC than do incongruent trials. In Experiment 1 (24 participants), the weight-morality embodiment effect was explored in the default metaphorical association condition (i.e., heavy-moral, light-immoral). In Experiment 2 (22 participants), the moral metaphor on the vertical dimension was activated before the task. Since it has been demonstrated that there were moral-up/immoral-down as well as light-up/heavy-down associations, the metaphorical association between weight sensation and moral concept might be changed into “light-up-moral /heavy-down-immoral”. By such manipulation, the modulation effect of the metaphorical association on the “weight-morality” embodiment was investigated. The results in Experiment 1 showed that the response time (RT) in the "heavy-moral, light-immoral" condition was significantly shorter, accompanied by a larger LPC than that in the "light-moral, heavy-immoral" condition. In Experiment 2, on account that the moral metaphor on vertical dimension was activated, the results showed a reversed pattern in relation to Experiment 1. Specifically, we found that the RT in the "light-moral, heavy-immoral" condition was significantly shorter than that in the "heavy-moral, light-immoral" condition, which was also accompanied by a larger LPC. Combined evidence in the two experiments indicated that the “weight-morality” embodiment effect was influenced by the metaphorical association. The larger LPC in the conditions congruent with the activated “weight-morality” association indicated that compared with the incongruent conditions, the “weight-morality” metaphorical association in the congruent condition was more salient and had a higher arousal level, which might promote the categorization response. In conclusion, results in this study indicated that the experience of weight could facilitate the processing of moral concepts, benefited from the metaphorical association of weight-morality, which guides and moderates the embodiment effect. This study suggested that the performance of the embodiment effects depends not only on the sensory-motor experience but also on the metaphorical associations primed in the embodiment process.