Dongting Lake is an important ecological barrier along the Yangtze River and is also important for social and sustainable economic development. An ecosystem services assessment of Dongting Lake shows flood control and floodwater storage is its most important ecological function. However, we know little about the factors controlling the evolution of the Dongting Lake flood storage function. We especially lack an analysis of the mechanisms of interaction between the landscape structure and flood storage conditions. Any landscape pattern is the result of various combined natural and human factors at different spatial and temporal scales. At Dongting Lake, these factors have had a profound influence and determine the hydrological procession of the wetlands storage capacity of the lake. Exploring the relationship between the structure-procession-function of the Dongting Lake area by studying the landscape patterns with specific ecological functions is very important. This study, based on landscape ecology and hydrology, uses landscape data from the Land TM satellite imagery from the Dongting Lake area from 1980, 1990, 2000, and 2005. Combined with the hydrological measurement data for these four years, the landscape and hydrological data were analyzed by using gray connection analysis, to discuss the influence and mechanisms of the landscape pattern and structure on the lake's flood storage capacity. The results show the landscape pattern of Dongting Lake area changed significantly from 1980 to 2005. The paddy fields, dry land, and human settlement areas changed the most dramatically, while the beaches changed less and forests, grassland and water changed the least. Numbers of patches (NP) of each land type and shape index (LSI) had similar trends, showing a surge between 1980 and 1990, stabilizing from 1990 to 2000, and then decreasing from 2000 to 2005. The aggregation of the various land types and the largest patch index (LPI) changed in different ways. The aggregation index (AI) and LPI of the paddy fields, forest and water changed in similar ways. The value of these indices gradually dropped from 1980 to 2000, and then increased from 2000 to 2005. The AI of the dry lands, grasslands, beaches and villages, and other types have contrasting changes when compared with the paddy fields; these increased from 1980 to 2000 and then decreased from 2000 to 2005. Flood storage capacity of the Dongting Lake area decreased in the first part of the study period and then increased later. This corresponded to the landscape pattern changes. Gray connection analysis showed there was clear relationship between the landscape patterns and flood storage capacity. The correlation was largest between the flood retention amount and the indices of LPI and AI, with correlation values of 0.77 and 0.75, respectively. This shows the obvious relationship between dominant patch size and the aggregation of homogeneous patches with flood storage capacity. The correlation is also pronounced between the silt deposition ratio and the NP and LSI indices, whose correlation values were 0.7 and 0.78, respectively. This illustrates the positive correlation between the silt deposition ratio and the scale of landscape fragmentation, diversity and shape complexity. The correlation was small between the landscape pattern indices and the flow diversion and sediments diversion ratios. The landscape indices and storage capacity correlation analysis revealed a high correlation between the flood retention capacity and the LPI and AI indices of paddy fields, woodlands, and water area. In conclusion, an increase of the LPI and AI of the paddy fields, woodlands, and water area would be helpful in increasing the flood storage function of Dongting Lake.