Different life-form aquatic vegetation support critical ecosystem services in most shallow lakes by improving water environment, and sequestering carbon and nutrients. Accurate spatio-temporal distribution dynamics of aquatic vegetation is basis to comprehensively obtain the changing trend of lacustrine ecological environment. In this study, we first constructed a high-precision machine learning algorithm to identify different life-form aquatic vegetation. Then we applied this approach to achieve the spatio-temporal distribution dynamics of different life-form lacustrine aquatic vegetation in the middle and lower reaches of the Yangtze River Basin. Our study showed that the overall classification accuracy was 0.81, and the Kappa coefficient was 0.74. The total area of aquatic vegetation ranged from 2541.58 km² to 4571.42 km² during 1986-2020, covering 15.99%-28.77% of the total lake area, and submerged vegetation was the dominated types (Max_1995yr=2649.21 km², Min_2005yr=921.38 km²), followed by emergent vegetation (Max_2005yr=1779.44 km², Min_2020yr=569.05 km²) and floating vegetation (Max_2015yr=685.68 km², Min_2000yr=293.04 km²). Aquatic vegetations were mostly located in the lakes of mainstream of the Yangtze River, and followed by Lake Poyang Basin, Lake Dongting Basin, Lake Taihu Basin and Han River Basin. It indicated that the area of aquatic vegetation showed a trend of increase (1986-1995), then decrease (1995-2010) and increase again (2010-2020). This study could provide a scientific basis for lake ecological research and management.