The many Amazon floodplain lakes are biogeochemical hotspots that have significant impacts on global biogeochemical cycles. As the Amazon seasonally inundates them and bring in nutrients and sediments, the lake-integrated river water fraction (CL) is a primary control of biogeochemical reactions in these lakes. Field studies were limited in their capability to thoroughly demonstrate the spatial and temporal distribution of river water. Here we used a catchment-lake processes model, validated by field study, to elucidate the lake mixing dynamics for a floodplain lake with a basin to lake area ratio (BA:LA) of 10.8. The inundation cycle can be divided into four phases: low water, rapid fill, co-rise, and receding. We show that in the co-rise phase, both catchment contributions and river stage are important and CL is at a close balance. CL has peaked long before the river stage peak, when local catchment water has pushed a significant portion of the river water out of the lake. Numerical experiments and variance partitioning show that basin size and seasonal rainfall distribution, e.g., rainfall in 5 leading months, have important controls on CL. However, their impacts vary by phases, as the river water and local dominate the rapid fill and low water phases, respectively. The basin area is thus more influential during the co-rise phase. Our results suggest modeling biogeochemistry in the Amazon floodplain lakes requires sufficient understanding of catchment hydrology and lake hydrodynamics.