In the last twenty years, several major terror attacks conducted in the name of political Islam hit Western Europe. We examine the impact of such terror attacks on hostile behavior on social media from a cross-national perspective. To this end, we draw upon time-stamped, behavioral data from YouTube and focus on the frequency and popularity (“likes”) of ethnically insulting comments among a corpus of approximately one hundred thousand comments. We study aggregate change and use individual-level panel data to investigate within-user change in ethnic insulting in periods leading up to and following major terror events in Germany, France and the UK. Results indicate that terror attacks boost interest in immigration-related topics in general, and lead to a disproportional increase in hate speech in particular. Moreover, we find that attack effects spill over to other countries in several, but not all, instances. Deeper analyses suggest, however, that this pattern is mainly driven by changes in the composition of users and not by changing behavior of individual users. That is, a surge in ethnic insulting comes from hateful users newly entering online discussions, rather than previous users becoming more hateful following an attack.