The present study was undertaken to examine the impact of Internet addiction on students’ social skills in smart and regular schools. The population of the study included the entire Tehrani junior and senior secondary school students in 2014-2015 academic year. The sample size was determined to be 384 based on Cochran’s formula and this number was increased to 400 students. For the purpose of sampling multi-stage cluster sampling method was used. The instrumentation comprised the Internet Addiction Test (Young, 1998) and social skills questionnaire. The obtained data were analyzed by using MANOVA. Results showed that there was no significant difference between smart and regular schools in terms of students’ social skills. Furthermore, in both smart and regular schools girls had the edge over boys regarding social skills. In addition, findings indicated that there was a significant difference between Internet-dependent students and those who did not show dependency on the Internet. In fact, Internet-dependent students demonstrated higher social skills.