The use of deconstructive approaches, such as functional classification models, can be an important tool for understanding aquatic systems. This study aimed to evaluate the distribution and community structure of mixotrophic flagellates in lakes of four large Brazilian floodplain systems, those of Amazon River, Araguaia River, Pantanal and upper Paraná River, in contrasting hydrological periods. A partial redundancy analysis (pRDA) was used to estimate the relative roles of environmental, spatial and biogeographic factors on both the entire community and a deconstructive approach using Reynolds' mixotrophic flagellate functional groups (RFGs). High environmental variability was found between the rainy and dry periods, mostly for the Amazon floodplain system. We identified 135 taxa and seven RFGs of mixotrophic flagellates: E, Lo, Q, W1, W2, Ws and Y. A permanova evidenced that biomass of mixotrophic flagellate RFGs reflected differences among floodplains and between hydrological periods. Our results showed that the use of RFGs better summarises the mixotrophic phytoplankton metacommunity structure when compared to the use of the entire community. The pRDA showed that explanatory matrices explained up to 37% of variation in the community. Environmental, spatial and biogeographic factors were important to explain the structure of the mixotrophic flagellates; however, the importance of the different components seemed to depend on the hydrological period and the traits of each RFG.