A MATLAB-based program was developed to simulate the distributions of states and behaviors of microbial storage product-accumulating bacteria in suspended growth systems. Currently available computer simulators of these systems predict dynamic behaviors by numerically solving differential biokinetic equations using average, or "lumped" system states (e.g., average microbial storage products concentrations). However, individual bacterial states are expected to diverge from average values, in part because individuals can have different hydrodynamic histories in terms of their residence times in upstream completely mixed flow reactors. The distributed state simulation program presented in this paper (DisSimulator 1.0) tracks individual bacteria as they move through a completely mixed reactor system. The program was evaluated for competition between polyphosphate-accumulating organisms (PAOs) and non-polyphosphate-accumulating heterotrophs in an enhanced biological phosphorus removal (EBPR) system for treatment of domestic wastewater. For identical systems and process conditions, simulations accounting for distributed states predicted larger anaerobic and aerobic solids residence time requirements for successful EBPR than did simulations using the lumped approach. One reason for this was that distributed simulations predicted large numbers of the PAOs were essentially inactive due to depleted or maximized storage product contents, while the lumped simulations predicted homogenous, 100% active PAO populations. Characteristic state profile shapes developed rapidly and were stable as total population numbers changed. Lumped state assumptions were demonstrated to produce large errors in predictions of EBPR system performance, and so consideration of distributed states may improve the accuracy of microbial storage products-based process simulations in systems with completely mixed hydrodynamics. (c) 2005 Wiley Periodicals, Inc.