Military theoretical considerations suggest that a networked command and control architecture will provide a more effective form of command and control under complex operations that demand a high tempo of action. This article presents an experimental study with the purpose to examine team performance under different conditions of command and control architectures and their resilience to complexity. The experiment was performed with the task to extinguish simulated fires in a microworld. Three factors were varied in the experiment: command and control architecture, the number of simulated units, and tempo. The dependent variable was the number of lost cells in the microworld. Three command and control architectures were investigated; command by negation, directive command, and a control condition. The general conclusion from this experiment was that all command and control architectures performed equally poorly under the condition of many subordinate units and fast tempo. This was in contradiction to suggestions made in the military theoretical literature. Command by negation was presumably the more effective command and control architecture under the other conditions.