Wetlands, unique ecosystems formed by the interaction of land and water, are susceptible to climate change and human activity. Landscape developmental processes are an important subject in the fields of geography and ecology. Landscape pattern index is an indicator of landscape structure, composition, and spatial configuration. In this paper, we discuss changes in landscape patterns in Baiyangdian wetland during 1984-2014, and the driving forces behind these changes, based on Landsat Thematic Mapper (TM) and GF-1 remote sensing images. A landscape classification system was formulated, and landscape pattern indices were selected at the landscape and class levels to reflect patterns in structural composition and spatial configuration and to analyze the evolution of the landscape. Our results showed that the area of emergent and submerged plants had decreased, and farmland and residential areas increased rapidly. Open water showed an "increasing-decreasing-increasing" tendency, while forest and bare land area did not change significantly in the study period. The wetland area (emergent plants + submerged plants + open water) decreased from 1984 to 2014, with an average area of 25008 hm² before 1998 and 21573 hm² after 1998. Emergent plants were the main cover type in Baiyangdian, covering 37%-61% of the total study area, followed by farmland, submerged plants, and open water. Interconversion among emergent plants, submerged plants, and open water occurred with changes in the hydrologic regime. Over the past three decades, farmland has had the largest average patch area, followed by emergent plants. Emergent plants had the highest patch area index and dimension index, indicating that this was a dominant cover type in the study area. In addition, the cohesion indices of farmland and emergent plants were highest and showed a smoothing trend, while residential areas, forestland, and bare land showed a discrete distribution, high degree of fragmentation, and inferior connectivity. Shannon's diversity indices showed a decreasing trend in Baiyangdian from 1989 to 2004, during which the landscape pattern tended to be unstable, and cohesion indices and wetland connectivity increased. Shannon's diversity indices, cohesion indices, and landscape heterogeneity increased between 1984 and 1989 and between 2004 and 2014. Population and economic development are major factors affecting changes in landscape patterns in Baiyangdian.