Computer simulation is used to investigate the success of an “outside agitator,” specifically the likelihood of a persistent extremist leading into “radical” action a gathering of moderates not otherwise disposed to militancy. The theoretical framework and computer simulation translation on which this work is based (Johnson&Feinberg forthcoming) assumes an ambiguous situation confronting a gathering in which each member can be located on a 10-point scale of action-choices. In the ensuing milling process, cues for action are offered and crowd members respond by shifting their choices and/or physical locations; thus, consensus for action is achieved and crowd structure emerges. In this test we generated a gathering of moderates and a persistently radical influence source, and varied whether gatherings initially maintained physical distance between themselves and the agitators. Results generally suggest that outside agitators usually cannot influence a crowd of moderates to take radical action. The variables most related to agitator success are crowd suggestibility, the action-choice advocated by the agitator, and the probability of movement within the physical space. Thus, the agitator is likely to be successful only in the specific and infrequent circumstances of a small gathering in a highly ambiguous situation in which the crowd members are not suspicious of the outsider. Outcomes showing the effects of microstructural conditions for emergent structure are also presented and discussed.