© 2017 Elsevier Ltd The purpose of this paper is to examine the impact of terrorism on tourism demand in Greece using monthly data from 1977 to 2012. We investigate whether this relationship is bidirectional and whether it exhibits long run persistence. Thus, we employ a large dataset of terrorist incidents and perform cointegration and long-run causality tests, correcting our data for cyclical seasonality and applying PCA to construct a terrorism proxy according to the severity of the incident. Our findings concur that terrorism has a significant negative impact on tourist arrivals to Greece and that causality is noted from terrorism to tourism only. The results suggest that authorities should establish firm measures against terrorism and that further actions should be taken to promote tourism, safety and security, as a response to terrorist incidents. Our study is, to the best our knowledge, the first to approach terrorism using a three-factor proxy with qualitative features.