Voting Advice Applications (VAAs) have become an integrated part of modern elections. While important debates remain about how to best construct VAAs(Germann et al., 2015; Holleman et al., 2016; Mendez, 2012), their documented effects on voter turnout and party choice brings important questions to the fore. From perspectives of political inequality and representation, it becomes important to assess whom the uses of VAA’s are. Early contributions to the literature found that VAA users are younger, more often male, more educated and more interested in politics than both the general population and the population of internet users (Boogers and Voerman, 2003; Garzia and Marschall, 2012; Hooghe and Teepe, 2007; Marschall, 2008). These findings are replicated in later studies of more recent elections and recently introduced VAAs (Dumont and Kies, 2012; Marschall, 2014; Ruusuvirta, 2010; Wall et al., 2009) This paper assesses whether the composition of users changes over time using German election data. The German case is instructive, because Germany has the most successful VAA measured in the absolute users. In doing so, the article assess whether VAA users over time becomes more similar to the population at large – something, which has been suggested but not tested in the literature(Marschall, 2014, p. 102). The article assess this normalization hypothesis by comparing age, educational level and degree of political interest of VAA users between different German election studies. The article assess the development from 2009 to 2017.