Survey oriented research about politics has used similar kind of items for decades. These survey items have allowed researchers to advance in the study of public opinion about politics in general, and in the study of gender differences in political attitudes in particular. Yet, some of the most commonly items measuring political attitudes were designed during the sixties in the context of the “behavioural revolution” in the social sciences in the US. Almost five decades later the political context has completely changed. The internationalization of politics, the emergence of new media, and new technologies have increased the availability of information, which has now become almost unlimited. Similarly, political actions once considered ‘unconventional’ have become habitual repertoires of political action. The position of women is now also different as it was fifty years ago, even if survey data show persistently that gender differences look quite the same as they did in the sixties. So far, however, little effort has been done to extend classic indicators of political attitudes beyond the restricted domain of party politics – to other aspects which might be of more interest for women. A critical question is indeed: What if gender differences are artificially constructed through survey items based on a restricted notion of politics? In this paper we present the results of an explorative on line survey to a representative sample of the Spanish population who has been asked what they think is political and what they think is not. We have specifically designed the survey under the expectation that there are significant differences in the way citizens understand and define politics. Indeed, our findings indicate that men and women have different conceptions of politics, which should be taken into account when designing survey instruments.