Behaviours associated with drug action can sometimes be elicited, in the absence of drug, by exposure to stimuli that were present during drug administration. Such a finding is usually interpreted as a conditioned drug effect. Often, however, the outcome could arise if drug administration in a particular environment retarded behavioural habituation to that environment. To test the 'habituation hypothesis' of conditioned drug effects, mice received 10 daily injections of damphetamine ('paired' group) or saline ('unpaired') in test boxes, and the converse injections in the colony room. Another group received saline in both environments. The apparatus and procedures yielded minimal habituation of behaviours (ambulation and rearing) over sessions. Only the paired group demonstrated behavioural sensitization, indicating environment-specific sensitization. The paired group also showed more ambulation and rearing than the others on the critical test of conditioning (saline injection in test box); moreover, their conditioning test scores were higher than those of the other groups on their first exposure to the test boxes, contradicting the habituation hypothesis. Further supporting the involvement of Pavlovian conditioning, levels of ambulation and rearing measured for 10min before each injection increased in the paired group, relative to the unpaired groups, over successive pairing sessions. Tests controlling for differential handling/injection experience produced results consistent with those previously obtained. Together, the findings are incompatible with the habituation hypothesis, and further support the role of Pavlovian conditioning.