VAA represents a recent tool that should work as shortcut for citizens willing to cast a reasonable vote. Although there are still doubt regarding its accuracy and reliability, it seems that the usage of this application can actually have positive impact on political behavior, in terms of turnout and in terms of motivation to gain more information (Garzia 2015). This finding could lead us to the conclusion that, despite the lack of cognitive abilities that was attributed to citizens (Converse 1964), nowadays VAA can help voters more than heuristics used to do (Gigerenzer et al. 1999; 2002). However, we are not considering who is actually interested in using this application. My insight is that a pre-selection in the use of VAA can in fact nullify the positive effect on political behavior. Therefore, it is important to understand to what extent people using VAA are those already informed and interested in politics. In doing so, I will use the PartiRep 2014 dataset to address this research question: does sophistication work as incentive for the usage of VAA? To measure sophistication, I will follow the indications of Luskin (1987), who proposed an index of size (factual knowledge), range (width of knowledge) and constraint (consistency of knowledge) of political cognitions at the individual level. The presence in PartiRep of questions on VAA’s usage and on the party suggested by the application will allow me to test my above-mentioned hypothesis and will possibly open new scenarios for the studies of Voting Advice Applications.