Western media and analysts increasingly speculate that civil unrest in Syria, Iraq, and Libya might lead to a redrawing of the map of the Middle East and North Africa. These accounts often claim or assume that the states and borders of the region have remained fairly constant since World War I. In this paper, I introduce a geographic information system (GIS) dataset of borders and border changes in the MENA since 1900 and demonstrate that the map changed considerably throughout the 20th century, although more so in some decades than in others. Depending on criteria, between 14 and 39 autonomous territorial polities disappeared (e.g., the Kingdom of the Hijaz, Hatay State, the Republic of Mahabad), most often by being absorbed into or conquered by other states. Other interstate borders moved, either through conquest or negotiated exchanges of land. I use my GIS to identify, classify, and visualize borders. I then ask what characteristics make some borders more durable than others. Are borders drawn by colonial officials without regard to geographic features less durable than those that run along earlier Ottoman administrative boundaries? Are boundaries that divide linguistic, tribal, and sectarian group more or less durable than those that do not? I explore several case studies of border disputes to understand how the characteristics of borders shape policy-makers’ willingness to change them. Did the seeming arbitrariness of the Jordanian-Saudi Arabian frontier (sometimes referred to as “Winston’s Hiccup”) make it easier for two kings to swap 5000 square miles in 1965? How did linguistic and religious divisions shape Turkish, Arab, and French understandings of what should happen to the Sanjak of Alexandretta? The paper concludes by revisiting the claim that the borders and states of the Middle East are artificial and amenable colonial constructs.