Under the dynamics of urbanization and counterurbanization, rural areas in China face both challenges and opportunities with the government’s new-town policy. A comprehensive analysis is essential for developing effective strategies. Characterized by traditional water settlements and high ecological sensitivity, Anxin County, which encompasses 50% of Baiyangdian Lake, was selected as a case study. Anxin County was incorporated into the National New Area in 2017, and it served as an experimental site for green ecological development. This study aims to provide insights into the sustainable spatial planning of Anxin County’s settlements by examining their long-term evolution and locational differences, modeling the driving mechanisms, and proposing differentiated spatial planning strategies based on predictive outcomes. Our research findings indicate the folllowing: (1) Anxin County’s settlements have expanded significantly in the past, with a notable surge between 1975 and 1996. Initially, semi-waterside settlements experienced the fastest growth before the 1970s, followed by land settlements. (2) Natural and socio-economic factors are modeled as independent variables to explain the evolution of settlement areas. The results indicate a decreasing impact of natural factors and an increasing influence of socio-economic factors over time. Furthermore, the evolution of settlements in water areas is relatively straightforward and random, whereas land settlements are influenced by a complex array of factors. (3) Utilizing the model to predict settlement growth, the study identifies settlements requiring relocation, and it proposes the most suitable relocation targets for them. A total of 23 villages, including Bei Tianzhuang and Cai Putai, are identified for relocation, while villages such as Ma Village, Bian Village, and Liu Lizhuang have strong capacities for accepting in-migrants. The study also offers recommendations for enhancing waterfront landscapes, flexible land use, road network systems, and internal construction.