The aims of the study were to test whether drug dose is positively related to the magnitude of the conditioned response following sensitization to the behavioural effects of cocaine and to investigate the relationship between the conditioned response and cocaine-induced sensitization. Male mice (C57BL/6J) were first injected over seven successive days with either saline or cocaine at 2.5, 5, 10 or 20 mg/kg s.c., in the testing room. On the test day, 24 h after the last injection, mice from all conditions were challenged with saline in the testing room to test for conditioned cocaine effects. Mice were video-recorded and various behaviours were later scored using a time-sampling technique. Cocaine-elicited orofacial stereotypy was significantly sensitized at the two highest doses and dose-dependently conditioned at the three highest doses. Cocaine-increased locomotion was sensitized at the three highest doses and significantly conditioned at 10 and 20mg/kg. Cocaine-increased sniffing did not change over pretreatment at any dose, and was conditioned only at 10 mg/kg. Cocaine-decreased immobility also did not change over pretreatment at any dose, but was conditioned at 10 and 20mg/kg. Concomitantly, rearing was reduced by cocaine at 10 and 20mg/kg, without sensitization being induced, and it was reduced under saline challenge after 5 mg/kg cocaine, while cocaine-decreased grooming was sensitized at the three highest doses and conditioned at 10 and 20 mg/kg cocaine. There was a positive relation between the size of the conditioned response for orofacial stereotypy and the magnitude of the unconditioned stimulus (the doses), a result conforming to the Pavlovian account of the placebo effect. This could also be concluded from considering the behaviour patterns as components of a unique placebo effect (hyperkinetic syndrome), since orofacial stereotypy, very apparent at 20 mg/kg cocaine, interfered at that dose with the full-blown expression of locomotion and sniffing, both yielding (approximately) inverted U-shaped dose-effect curves. However, no correlation was found between the magnitude of the conditioned response and the amplitude of sensitization (the difference between the initial unconditioned non-sensitized response and the last unconditioned sensitized response), a finding which indicates that conditioned responding does not participate in the generation of the sensitized effects, contrary to the 'excitatory conditioning model of contextual sensitization'.