Does the decline in party identification lead to a decrease or an increase in affective polarization? In recent years, research about affective polarization has increased, asking whether contemporary publics polarize in terms of their affective evaluations of the opposite party. Evidence shows that, at least in some cases, there are signs of increased polarization. At the same time, however, there is evidence of a decline in party identification, suggesting that the parties no longer attract people's hearts and minds. These two results might conflict. However, whether and how affective polarization and declining partisanship are related has received little attention. To address this issue, in this article, we investigate how much affective polarization there is in Chile, how it has changed over time. We use survey data from Chile between 1990 and 2021, a country that has shown a profound and constant loss in partisanship. First, we show that affective polarization varies over time and that, at the aggregate level, the decline in partisanship does not impact affective polarization. Second, the groups that show higher polarization also change: if by 1990 the more polarized were people identifying with left-wing parties, by 2021, affective polarization is similar across groups, including those who do not identify with political parties.