Membrane filtration treats drinking water by physical removal of bacteria and other particles present in the raw water. In order to study post-filtration contamination and growth, filtered river water and wastewater were used in a controlled laboratory-scale simulation of a batch-operated membrane filtration system. Bacterial batch growth was analyzed following intentional initial contamination with a river water microbial community. Batch growth in the permeate was measured with online flow cytometry at high intervals during 10 successive 24-hour operational cycles, simulating repeated daily use (filtration followed by stagnation). Two operational mechanisms influenced the growth characteristics: (1) initial selection of bacteria adapted to batch growth conditions, and (2) biofilm formation on the surfaces of the permeate containers. The first mechanism contributed towards a stable and reproducible growth behavior (lag phase of less than 4 h, maximum growth rates of 0.37-0.42 h(-1) and final total cell counts of 1.5-1.8 x 10(6) cells mL(-1)) throughout several consecutive operational cycles. When the feed water changed, the adapted bacterial communities grew rapidly and proportionally to the amount of substrate in the new water source. Biofilm development in the permeate containers resulted in a 20% reduction in the overall cell production during five operational days, suggesting this to be a potential novel strategy towards controlling biological stability in such systems. (C) 2015 Elsevier B.V. All rights reserved.