Water level fluctuations resulting from natural and anthropogenic factors have been projected to affect the functions and structures of wetland vegetation communities. Therefore, it is important to assess the impact of the hydrological gradient on wetland vegetation. This paper presents a case study on the Honghe National Nature Reserve (HNNR) in the Sanjiang Plain, located in Northeast China. In this study, 210 plots from 18 sampling line transects were sampled in 2011, 2012, and 2014 along the hydrological gradient. Using a Gaussian logistic regression model, we determined a relationship between three wetland plant species and a hydrologic indicator—a combination of the water level and soil moisture—and then applied that relationship to simulate the distribution of plants across a larger landscape by the geographic information system (GIS). The results show that the optimum ecological amplitude of Calamagrostis angustifolia to the hydrological gradient based on the probability of occurrence model was [0.09, 0.41], that of Carex lasiocarpa was [0.35, 0.57], and that of Carex pseudocuraica was [0.49, 0.77]. The optimum of Calamagrostis angustifolia was 0.25, Carex lasiocarpa was 0.46, and Carex pseudocuraica was 0.63. Spatial distribution probability maps were generated, as were maps detailing the distribution of the most suitable habitats for wetland vegetation species. Finally, the model simulation results were verified, showing that this approach can be employed to provide an accurate simulation of the spatial distribution pattern of wetland vegetation communities. Importantly, this study suggests that it may be possible to predict the spatial distribution of different species from the hydrological gradient.