The in ux of immigrants to Norway over the last decades is a large-scale natural experiment. This paper exploits municipal-level variations in the immigrant population (1977 2011) to estimate the causal e¤ects on voter support for the right-wing, anti-immigration Progress Party. The results indicate that voters keep incumbents accountable for permis- sive immigration policies. Immigration from non-Western countries (Africa, Asia, Latin America) has increased electoral support for the Progress Party. However, the e¤ects are quite modest and noticeable only in the initial phases of immigration. Survey data covering ten elections (1989 2011) in- dicate a similar development in anti-immigration attitudes. The primary immigration shock tends to burn out quite fast as people get direct experi- ence of immigrants on a daily basis.