Strong negative feelings towards political parties and their supporters seem to be on the rise and have been shown to have important implications for behaviour in- and outside of politics, such as political participation, dating behaviour and even job market discrimination. Most of the current literature on negative partisanship was developed in two-party systems where it is often presented as a by-product of positive partisanship. However, there is growing evidence that negative partisanship is also prevalent in multi-party systems; and that it is distinct from positive partisanship. In this paper, I explore experimentally what consequences negative partisanship has for political behaviour in multi-party systems. I argue that negative partisanship increases vote switching by changing how voters evaluate all parties on offer, not just the one they feel the strongest dislike towards. I test these ideas using a survey experiment conducted in Canada. I find that priming voters about negative partisanship has an effect on party evaluations, and that this effect is not just limited to those voters who also have high levels of positive partisanship. This suggests that negative partisanship has important implications for political behaviour, even in multi-party settings. Negative partisanship is not the mere by-product of positive party identification but has powerful consequences of its own.