Land use planning involves categorizing spatial units on the basis of land use type to yield optimum productivity. While the planning is known to have two approaches-land use proportion optimization and land use layout optimization,-studies have mostly focused on the former quantitative approach, rarely attending to the latter. The existing literature applies some time-consuming heuristic methods to optimize land use layout, and therefore, this study attempts to solve this using the CLUE-S model with a reference point approach. While this model has been widely used to simulate future land use layout, there exist great differences between optimization and simulation, as optimization aims to achieve better benefits from land use layout, while simulation is used to deduce a layout with rules involving historical changes. In order to realize the multiple objectives of land use and to optimize land use layout for spatial welfare, thus increasing land resource allocation efficiency, this study focuses on the regional land use layout optimization under different scenarios. First, a reference point programming approach is formulated to optimize land use structures in the different scenarios. Then, optimal distribution rules for different land use types are extracted from current land use map through logistic regression. Finally, using optimal land use structure as the quantity constraint, the CLUE-S model embedded with optimized layout rules is applied to optimize overall land use layout in the different scenarios. The results of this study show the following:the economic value of land use is 1040 billion RMB under the economic priority scenario, the ecological value is 9.7 billion RMB under the ecology priority scenario, and the area of land that represents human well-being is 351271 hectares under the scenario of public well-being priority. The goal of economic growth is the most sensitive to scenario variation of land use structure, while the public well-being goal is the most insensitive to scenario variation. Compared with the existing layout, there are drastic differences in the land use optimal layout under the economic priority scenario, where urban land expands towards east and northwest as well as in the ecology priority scenario, where the amount of garden and forest land was allocated to the hilly area in the southwest and the Yangtze River side in the south of downtown of Yangzhou city. The method of optimizing land use layout under different scenarios as proposed in this study could supply effective technical assistance for land use plan and ecological plan.