Political scientists have acknowledged the importance of ethno-nationalism as a constitutive element of radical-right politics, but have typically empirically reduced the phenomenon to its downstream attitudinal correlates. Sociologists, on the other hand, have extensively studied nationalism, but have rarely weighed in on debates about institutional politics. In this study, we bring these literatures together by considering how nationalist beliefs shaped respondents' voting preferences in the 2016 U.S. presidential election and how the election outcome built on long-term changes in the distribution of nationalism in the U.S. population. The results suggest that competing understandings of American nationhood were effectively mobilized by candidates from the two parties, both in the 2016 primaries and the general election. Furthermore, over the past twenty years, nationalism has become sorted by party, as Republican identifiers have become predominantly ethno-nationalist and Democrats have increasingly endorsed creedal and disengaged conceptions of nationhood. This points to the rising demand for radical candidates among Republicans and suggests a potentially bleak future for U.S. politics, as nationalism becomes yet another among multiple overlapping social and cultural cleavages that serve to reinforce deep partisan divisions and undermine the stability of liberal democratic institutions.