This descriptive scoping aims to understand how the prevalence of cyberbullying has been estimated across studies. A systematic scoping review of cyberbullying empirical studies was conducted by using three bibliographic databases to search for papers published between January 2004 and August 2014. A protocol was defined to identify the relevant papers. Papers selected were included in a data sheet developed by the authors to record specific findings. In total, 159 studies were included in the scoping review. Most of the prevalence studies were conducted in the last 4 years, mainly in North America (n = 77) and in Europe (n = 65). High methodological heterogeneity was found among the studies, which may contribute to explain variability in prevalence estimates. Cyberbullying experiences were assessed through several different perspectives: focused only on victims, focused only on perpetrators, or focused on both victims and perpetrators (without differentiating between if they are victims or perpetrators). Most of the studies tend to assess cybervictimization experiences. However, even considering the same perspective, the same country, and the same recall period, a high variability in the estimates was observed. As a main conclusion, the way in which the prevalence of cyberbullying is estimated is influenced by methodological research options.