Adolescents are known to engage in more risk-taking behaviors(such as smoking, drug use, skiing, and climbing) than children and adults. Previous research has suggested that at least two factors are associated with adolescent risk-taking behavior: one is the presence of a peer, which could activate the brain’s reward circuitry and lead to adolescents’ more reward preference and risk-taking; the other is self-esteem, high self-esteem individuals are inclined to take part in more risky behaviors, particularly those socially acceptable risky behaviors. However, the interaction between peer presence and self-esteem on adolescent risk-taking behavior is still unclear; especially evidence from ERP studies is rare. With ERP technique, the present study explored the interaction between peer presence and self-esteem on adolescent risk-taking behavior. According to the scores on Rosenberg’s Self-Esteem Scale, 34 participants were chosen from 216 freshmen and were divided into two groups: the 16 students with top scores were in high self-esteem group and the 18 adolescents with lowest scores were in low self-esteem group. Each group was then divided into two subgroups: one was peer presence subgroup in which each adolescent was asked to bring a sex-same peer to come to observe his or her following experiment performance on the Balloon Analogue Risk Task(BART); in the other subgroup without a peer in presence, all participants completed the BART alone. All participants’ EEGs were recorded when they were operating BART and their BART values(total number of inflating balloons/number of unexploded balloons) were also calculated as their behavioral index of risk-taking behavior. Behavioral results indicated that: 1) those adolescents with either a peer in presence or high self-esteem showed more risk-taking behaviors; 2) self-esteem moderated the effect of peer presence on adolescent risk-taking behavior: only for high self-esteem adolescents, did peer presence significantly increase their risk-taking behaviors. The ERP experiment showed similar results: 1) compared to the alone group, the group with a peer in presence demonstrated larger amplitudes of N1, P3 and LPP; and high self-esteem adolescents’ P3 and LPP amplitudes were also greater than those with low self-esteem, indicating possibly more risk-taking behaviors; 2) the effect of peer presence was only significant for high self-esteem adolescents who showed greater amplitudes of N1, P3 and LPP components when observed by a peer. These findings provide ERP evidence for the effects of peer presence and self-esteem on adolescent risk-taking behavior and suggest that adolescents with high self-esteem should be more influenced by their peers on their risk-taking, particularly on that non-negative risk-taking, possibly because they are more sensitive, more exciting, and have a stronger approach motivation to rewards in risk-taking.