The present study explores the language of evaluation in a sub-genre of political discourse, pre-electoral debates, and its potential persuasive function for gaining voters via a contraposition of positive self-evaluation and negative evaluation of the other candidate. A further aim of this research is to check whether the candidate¿s ideology has a bearing on the entities that get evaluated. After a brief examination of the characteristics of the sub-genre at hand, specifically in the Spanish context, we present the results of an evaluation analysis carried out in a corpus of 19,849 words, which is the extension of the most recent pre-electoral debate held in Spain between the candidates of the two main political parties. Taking into account Van Dijk¿s CDA framework (2005) for parliamentary debates as global semantic strategies of positive self-presentation and negative other-presentation, Martin and White¿s (2005) method was adopted as an analytical tool. The results showed that, although each candidate had different preferences in the choice of evaluative devices, they both used them as a strategy to win electoral votes while deprecating the opposing party and, therefore, minimizing their chances of winning the elections. On the other hand, and despite their opposing ideology, they both seem to defend those policies that are more widely accepted in order not to risk losing voters: public services and egalitarian social policies.