Recent political events have drawn greater attention, among scholars and among other political observers, to the possibility of turmoil and change in the party system. Yet, despite a remarkable depth of scholarship on American political parties, remarkably little is known about how parties in the United States emerge. Existing approaches to political science have offered reviews or sketches of party formation but have not treated it as a distinct research topic. This dissertation argues that successful party formation is caused by `strategic cadre-building,' a strategic orientation toward building a broad coalition of elites able to leverage their own resources toward victory for the broader party. Existing models of political parties depend on variables of party system crisis and appropriate ideology to conceive of the emergence of a new party. I show that these variables are not easily specified and do not appear to distinguish between new party success and failure. Rather, parties with similar ideologies succeed and fail within the same or very similar crises. Instead, successful instances of new party emergence are characterized by elites making the strategic choices that bring other elites into the party and thereby favor success. However, to do so, they build a party in which competition, for seats and for policies, exists. As a result, new party success often has a surprisingly limited impact on the political order. This project conducts a series of paired case studies comparing cases in which a new party successfully formed to similar cases in which a new party failed. In this way the Democratic Party is compared to the Jacksonians, the Whig Party to the National Republican Party, and the Republican Party is compared to the Free Soil Party. In each case, extensive qualitative historical analysis combines detailed evaluation of secondary sources with new review of the primary sources. In each case, the evidence strongly favors cadre-building as the determinative variable in new party formation.