A representative sample (n=1,000) of the Belgian population aged 18 years and older filled out an online questionnaire on their Internet use in general and their use of social networking sites (SNS) in particular. We measured total time spent on the Internet, time spent on SNS, number of SNS profiles, gender, age, schooling level, income, job occupation, and leisure activities, and we integrated several psychological scales such as the Quick Big Five and the Mastery Scale. Hierarchical multiple regression modeling shows that gender and age explain an important part of the compulsive SNS score (5%) as well as psychological scales (20%), but attitude toward school (additional 3%) and income (2.5%) also add to explained variance in predictive models of compulsive SNS use.