With its rapid diffusion, the Internet is playing a more and more important role in the daily lives of Chinese children. Literature shows the Internet is inconsistently associated with loneliness. The impacts of the Internet on children’s loneliness may lie in how children use the Internet. And personality could be an important moderating variable. The present study aimed to examine the impacts of types of Internet use (such as information seeking, online entertainment and online communication) on children’s loneliness and the moderating effect of children’s personality. Internet Use Questionnaire, Illinois Loneliness Questionnaire, Eysenck Personality Questionnaire (EPQ), and Family Affluence Scale (FAS) were administered to 637 Grade 3?6 children to examine the relationship between Internet use and children’s loneliness and the moderating effect of personality. The surveys were conducted in the participants’ classrooms and trained research assistants were present during all testing. All the completed questionnaires were collected by research assistants on the spot. No teacher was present during all testing to avoid response bias. Multiple regression analyses were employed to examine the impacts of Internet activities on children’s loneliness and the moderating effect of personality. Age and family income were entered as controlled variables. The results showed that: (1) children used the Internet more for information seeking and entertainment than communication. The most popular Internet activities for children were seeking information related to learning and life, online gaming, online video, and online music. Online communication was less prevalent in children. (2) High frequency of online information seeking significantly predicted children’s low loneliness（β= ?.20, p<.001）. (3) Personality moderated the relationship between some types of Internet use and children’s loneliness. First, social compensation effects were found in the relationship between online communication and loneliness. High frequency of online communication significantly predicted low loneliness only for those high neurotic children（β=?.20，p<.05）. While for those low neurotic children, the association was insignificant. Second, poor-get-poorer effects were found the relationship between online entertainment and loneliness. High frequency of online entertainment significantly predicted high loneliness for those introverted children（β=.24，p<0.05）. While for those extroverted children, the association was insignificant. Personality did not moderate the relationship between online information seeking and loneliness. The present study extends our understanding of the impacts of the Internet on loneliness. The results suggest that personality moderates the relationship between some Internet activities and children’s loneliness. The study has some practical implications for parents and teachers to help children use the Internet appropriately. Furthermore, almost all previous studies are concerned with adolescents and adults, leaving children’s Internet use unexplored. The present study bridges these gaps by focusing on children’s Internet use.