Swarming occurs when several units conduct a convergent attack on a target from multiple axes. In this dissertation the author uses case studies, comparative analysis, and common sense to derive a simple theory that explains the phenomenology of swarming. He researches 23 case studies of swarming, ranging from Scythian horse archers in the fourth century BC to Iraqi and Syrian paramilitaries in Baghdad in 2003, to understand swarm tactics and formations, the importance of pulsing, and the general characteristics of past swarms. He considers command and control, communications, home field advantage, surprise, fratricide, and training. The author identifies five primary variables most important to successful swarming: (1) superior situational awareness, (2) elusiveness, (3) standoff capability, (4) encirclement, and (5) simultaneity. Treating the five variables as binary — either absent or present in a case — he derives 32 possible combinations of these variables that together comprise a “model’ that predicts swarming outcomes based on his theory. He predicts that only six combinations lead to swarm success. The model is tested using a qualitative technique called the comparative method to find patterns of multiple and conjunctural causation. Finally, the author addresses the questions of how swarms can be defeated and whether swarming is relevant for future friendly forces.