As a part of the systems modeling research at NASA Ames Research Center, the use of a market-based control strategy to actively manage power for a model of a regenerative life support system (LSS) is examined. Individual subsystem control agents determine power demands and develop bids to ‘buy’ or to ‘sell’ power. A higher level controller collects the bids and power requests from the individual agents, monitors overall power usage, and manages surges or spikes. The higher level controller conducts an ‘auction’ to set a trading price and then allocates power to qualified subsystems. The auction occurs every twelve minutes within the simulated LSS. This market-based power reallocation cannot come at the expense of life support function. Therefore, participation in the auction is restricted to those processes that meet certain tolerance constraints. These tolerances represent acceptable limits within which system processes can be operated. The power manager itself enters the auction to facilitate trading as needed, while simultaneously mitigating or eliminating excessive power surges. We present a simulation model and discuss some of our results.