There seem to be no limits on what can pass through state constitutional amendment procedures. State amendments have targeted vulnerable minorities, deeply entrenched specific fiscal strategies, and profoundly restructured institutions. The malleability of state constitutions is significant because in many states there are legitimate fears that special interests dominate amendment politics, and that fundamental change is occurring with minimal opportunities for constructive deliberation or inclusive participation. The state doctrine of “referendum sovereignty” is a key condition fueling this dynamic. The doctrine holds that there are no substantive limits on any state amendment processes so long as amendments comply with federal law, explicit state procedural requirements, and are subject to a referendum. The doctrine assumes that a referendum is the full institutional embodiment of the people’s sovereignty, and it has led courts to uphold amendments that strip minorities of deeply embedded rights and even replace constitutions wholesale. This article provides the first systematic assessment of the doctrine of referendum sovereignty. To do this, it relies on a set of largely neglected sources: the debates of all known state constitutional conventions where state amendment processes were forged (ninety-one conventions from 1818 to 1984). These sources suggest that the underlying logic of state constitutionalism is inconsistent with the presumption of a limitless amendment power. Properly understood, state amendment processes are built on the assumption that only a constitutional convention of specially elected delegates is presumed to have inherent power to create or destroy a constitution. Extra-conventional amendment actors, on the other hand, are presumed to be subordinate creations of the existing constitution with the limited authority to propose modifications and without authority to destroy the constitution’s fundamentals. This distinction came from a deep distrust of existing officials and private groups. My findings are important in light of the ever-expanding use of extra-conventional amendment processes to effectuate large-scale change. They provide a coherent framework for state courts to assess whether a reform is an appropriate use of the amendment power or an unauthorized intrusion on the role of the convention.