Aggression can be costly to foragers, yet some recent research suggests that foragers should use aggression as a cue to patch quality (the attractive aggression hypothesis). If aggression is predictive of patch quality, then the attractive aggression hypothesis predicts that the distribution of foragers should follow the distribution of aggression. If, instead, aggression is repulsive because it is costly, then the distribution of foragers should diverge from the distribution of aggression. We tested the attractive aggression hypothesis using female guppies, Poecilia reticulata, and found that the distribution of foragers followed the distribution of food, but was unaffected by the distribution of aggression. These data do not support the attractive aggression hypothesis, but instead suggest that the distribution of aggression is a consequence of the distribution of foragers, and that aggression is not used as public information about patch quality.