Abstract: Social withdrawal is a broad construct with meaningful concrete indexes that featuring with shyness, behavioral inhibition, isolation, rejection, social reticence, passivity and peer neglect.. There are many subtypes under this big umbrella of social withdrawal. Social disinterest is a special subtype which refers to no particular interest in social interaction and expected to spend more time alone. They refrain from social interaction just because no strong motivation to join in the group and no strong motivation to averse to peer activities. The purpose of this study was to examine the relation between social disinterest and peer preference through a one year longitudinal study and to explore the possible causality by using cross-lagged analysis. The data was collected in September 2009 (Time1) and May 2010 (Time2). The procedure was the same at 2 times. 787 students in grade 4~8 were investigated by peer nomination and Revised Class Play (RCP). The results indicated that: (1) peer preference was negatively associated with social disinterest (r=-.43 at Time 1 and -.46 at Time 2, ps<.01); (2) both peer preference and social disinterest were highly stable within one year (r=.67 and .66 on peer preference and social disinterest, ps<.01). Repeated measure revealed that there was a significant three-way interaction between time, gender and grade (F（2, 782）=4.77, p<.01). Within 1 year, boys’ level of social disinterest in primary school was declined and girls’ was increased while boys’ level of social disinterest in junior high school was increased and girls’ was declined. (3) Cross-lagged analysis indicated that relation between social disinterest and peer preference was reciprocal on the whole. But gender differences were also found. For boys, Time1 social disinterest could predicted Time2 peer preference(β= -.14，p<.01) after controlling Time1 peer preference. However, Time1 peer preference could not predicted Time2 social disinterest (β= -.02, p>.05) after controlling Time1 social disinterest. For girls, an opposite pattern was found. Time1 social disinterest could not predicted Time2 peer preference (β=.05, p>.05) but Time1 peer preference could predicted Time2 social disinterest (β= -.23, p<.01) after controlling the Time1 stability. Therefore, the possible causality was different for boys and girls. The results may help us understand developmental relation of children’s social disinterest and peer preference in Chinese culture and some reasons for this gender-specific pattern were discussed.