Intense rainfall is one of the most serious and common natural events, causing the excessive inflow of rainwater into wastewater treatment plants. However, little is known about the impacts of rainwater dilution on the structure and function of the sludge microorganisms. Here, high-throughput sequencing of 16S ribosomal RNA (rRNA) genes was implemented to describe the microbial community dynamics during the simulated intense rainfall situation (event i) in which approximately 45 % of the sludge biomass was artificially overflowed by massive water supply in a pilot-scale membrane bioreactor. Thereafter, we investigated the functional and structural responses of the perturbed microbial communities to subsequent conditional changes, i.e., an increase in organic loading rate from 225 to 450 mg chemical oxygen demand (COD) l(-1) day(-1) (event ii) and an addition of a microbiota activator (event iii). Due to the event i, the COD removal declined to 78.2 %. This deterioration coincided with the decreased microbial diversity and the proliferation of the oligotrophic Aquabacterium sp. During the succeeding events ii and iii, the sludge biomass increased and the COD removal became higher (86.5-97.4 %). With the apparent recovery of the reactor performance, microbial communities became diversified and the compositions dynamically changed. Notably, various bacterial micropredators were highly enriched under the successive conditions, most likely being involved in the flexible reorganization of microbial communities. These results indicate that the activated sludge harbored functionally redundant microorganisms that were able to thrive and proliferate along with the conditional changes, thereby contributing to the functional maintenance of the membrane bioreactor.