Background: The number of left-behind children in rural China has increased dramatically over the last decade. It is reported that about 21.88% of child population with an estimated number of 61million are left-behind children whose parents leave them to work in cities. We conducted a cross-sectional study to explore the impacts of left-behind experience (LBE) on college students' depression and other influencing factors. Aim: This study discusses the mediation effect of self-esteem together with psychological resilience on college students with depression and negative life events of left-behind. The study also discusses the regulation effect of LBE. Methods: A total of 788 college students were selected from three universities in Sichuan and Chongqing (367 with LBEs, 421 without LBEs). Adolescent Self-Rating Life Events Check List (ASLEC), Self-Esteem Scale (SES), Resilience Scale of Chinese Adolescent (RSCA) and Self-Rating Depression Scale (SDS) were used to measure the negative life events, self-esteem, psychological resilience and depression, respectively. Bootstrap program was used to test the mediation effect, and multiple-group analysis was used to examine the regulation effect for LBE. Results: Scores of ASLEC for the college students with LBEs were higher than those without LBEs (8.593.57) vs (7.06 +/- 3.38), p<.001). The scores of LBE, ASLEC and SDS were positively correlated with the college students with LBEs (r=.21 to .29, p<.01), while the scores of RSCA and SES were negatively correlated (r=-.30 to -.59, p<.01). The mediation effect of college students' self-esteem and psychological resilience between negative life events and depression was significant (mediating effect=.08, .13, .07; p<.01). Thus, the college students' self-esteem and psychological resilience on negative life events had strong mediation effect on depression. The test of Bootstrap showed that the mediation effect of self-esteem and psychological resilience was significant (95% confidence interval (CI)=[0.04-0.76]). The LBE had regulation effect on college students' self-esteem and psychological resilience. (The constraint model fitting degree of variation is (2)=2,120.68, df=8, p<.001.) The self-esteem and psychological resilience of college students with LBEs can be used to mediate the relation between negative life events and depression, whereas those without LBEs cannot. Conclusion: Self-esteem and psychological resilience fully mediates college students' negative life events and depression, which is regulated by their previous LBEs.