This study investigates the relationship between students' use of information and communication technology (ICT) for social communication and their computer and information literacy (CIL) scores. It also examines whether gender and socioeconomic background moderates this relationship. We utilized student data from IEA's International Computer and Information Study (ICILS) to build multivariate regression models for answering the research questions, and accounted for the complex sample structure of the data by using weights for all statistical analyses, employing jackknife repeated replication for variance estimation. Students who frequently use the internet for messaging and participation in social networks (i.e., at least once a week) scored on average 44 points higher than those who use ICT for the same purpose only less than once a week or never. The direction of this effect was the same in all 21 participating educational systems, the difference ranging from 19 to 75 points (always statistically significant). We continued the analysis by testing whether the relationship is moderated by gender; as girls use more often ICT for social communication and have higher CIL scores on average. After controlling for the gender effect the CIL scores between the two examined groups decreased only by 2 points on average. Even after including students' socio-economic background into the model, the difference in CIL between the two groups of interest declined only little-to 32 points on average across all countries. The difference remained to be statistically significant in all countries but one. The results suggest a strong relationship between students' CIL proficiency level and the frequency of their use of electronic devices for social communication; hence, respective skills needed at schools and later on at the workplace are reflected in their use outside of school and for socializing.