The question of how and when voters punish corrupt politicians remains central in political science. To understand why and under what conditions people may overlook or minimally punish corrupt politicians, recent survey studies have identified various mitigating factors. However, there remains limited insight into when different types of corruption are penalized, especially since survey experiments often fall short of explaining tolerance for corruption. Our paper offers a novel investigation into Hungary, deemed the most corrupt country in Europe by Transparency International. Using a conjoint survey experiment conducted shortly before the June 2024 elections, we analyze how voters assess and prioritize various candidate attributes, including corruption; the conditions under which they decide to punish or tolerate corruption; and their tolerance thresholds for corrupt behavior. Additionally, by leveraging the unique context of the June 2024 elections—when, for the first time, both EP and local elections were held on the same day—we test how the influence of mitigating factors in shaping voters’ reactions to corruption varies between local and EP elections. Preliminary findings reveal two key results. First, voters are more tolerant of corruption when the act is perceived as less severe, when it brings clear side benefits, or when the information comes from a non-credible source. Second, we show how voter reactions to corruption vary across electoral contexts: in local elections, corruption is punished more severely, whereas European elections elicit a weaker response to corruption, as they are perceived as less directly tied to local issues.