AbstractThe behavioral response of established colonies of domesticated rats to the presence of an unfamiliar intruder of the same species represents one of the most effective procedures yet developed to study aggression in the laboratory. Here, the social, experiential, and environmental variables that influence attack severity are reviewed and several important methodological issues are discussed. Brief exposures of intruders to intact colonies may produce misleading results but long‐term test sessions increase the likelihood that intruders will be either killed or severely injured. We describe a simple modification of the colony‐intruder procedure whereby intruders can successfully defend themselves during long sessions and thus reduce serious injury. The modified procedure appears to conform more closely to what happens during aggressive encounters in free‐living populations of wild