The origins, development, and makeup of early state societies in China have long been a favorite topic of research, though there has recently been an upsurge of attention among archaeologists in China and abroad. Research has been dominated by the identification of the Erlitou site from the early second millennium BC as the center of the earliest state in China, sometimes identified with the Xia Dynasty. Recently, several scholars have employed neo-evolutionary criteria for the identification of Erlitou society as China’s earliest state in an attempt to provide objective criteria for the traditional historiographical narrative. Overarching social and ecological models of cultural change have been severely criticized by anthropological archaeologists, and many archaeologists studying the development of ancient societies prefer to focus on individual case studies or specific institutions rather than on the state. In contrast to recent archaeological scholarship that has called for its total abandonment, we find the “state” a useful concept for understanding local trajectories as well as cross-cultural comparisons. In this article we suggest a way of incorporating the warnings against simplistic overarching models while maintaining the notion of rapid sociopolitical change associated with state formation. Based on an analysis of the long-term trajectory, we identify, in north China, two phases of rapid transformations: the first, starting around 2500 BC, when several unstable regional states evolved and declined, and the second, around 1600 BC, when an intraregional state, usually identified with the historical Shang, rapidly evolved.