Real-life identities may differ from “virtual” identities presented in social networks. Many social networkers construct diverse (more than one) virtual identities, which differ by numerous parameters. Two or more identities constructed by a social networker we call alternative identities. Empirical study is based on a case developed by the Harvard Good Play Project; semi-structured interviews were held with 42 participants (social networkers) in three age groups (15—17, 18—21, 22—25; equal number of females and males in each group). Computer content analysis, narrative analysis of the interview materials and variance analysis were done. Results refer to reasons for and attitudes towards construction of alternative identities; age and gender differences are analyzed.