We trapped two populations of Niviventer confucianus in the Thousand Island Lake region using the capture-mark-recapture (CMR) method from July to November 2009 and from March to September 2010, and explored the dispersal characteristics of the N. confucianus populations. The Biotas 2.0 software was used to calculate the dispersal distance. The X² test was used to analyze the effects of sex, season, and age on the number of dispersers. One-way analysis of variance (ANOVA) was used to analyze the seasonal differences in body weight of the dispersers and the effects of age on the dispersal distance. The t test for independent samples was used to analyze the body weight differences between dispersers on two islands in the same season and the effect of sex on the dispersal distance. Linear regression analysis was used for correlation analysis. Our results showed that the dispersal ratio of N. confucianus was 37.96%, which is much higher than that of the Meriones unguiculatus and Microtus fortis that lived in the mainland. We considered the possibility that food and space limitations on the islands may lead to a dispersal ratio that is higher than that of mainland populations. The dispersal ratio showed significant differences among seasons, and the dispersal peak occurred in the spring, especially in March. Other studies showed that the peak breeding of N. confucianus occurs in April and May. We believe that the breeding behavior was fully active in the spring, contributing to the larger population size, and the breeding pressure led to the dispersal of some individuals. Sex-biased dispersal was detected, and the males were more likely to disperse (X² = 12.542,df = 1,P < 0.001). We assumed that more males dispersed in order to avoid inbreeding or to get more breeding opportunities, but this notion still needs genetic verification. A significant difference was found in the breeding period (X² = 4.504,df = 1,P = 0.034) when the pregnant rats were counted, which means that the pregnant females affected the dispersal model during the breeding period. Compared to the immature males at the subadult stage, sexually mature males were likely to disperse. The body weight showed no significant difference between the two sexes. The number of dispersers was related to their age, and more dispersers were found in the subadult group than in the other groups. The juvenile group was limited by their capacity for activity, and the adult group had enough food and spousal resources for their dominant positions in the population, but the subadult group exerted stronger pressure on food resources and reproductive opportunities. There was a significantly positive correlation between the number of dispersers and the size of social groups: the larger social group contained more dispersers. Social-group size reflected the level of competitive pressure on spouse resources, which meant that the social pressure may be the reason for dispersal of N. confucianus. The dispersal distance showed a highly significant difference between males and females: the dispersal distance of females ((282.97±16.14) m) was greater than that of males ((146.70±6.80) m). This phenomenon may be caused by the promiscuity system of N. confucianus or avoidance of inbreeding. Female offspring tended to disperse farther away from the maternal nest to claim a new territory in order to avoid resource competition, while reducing the inbreeding probability. The dispersal distance showed no significant relation with age, body weight, and the maximum movement distance.