The effect of confinement during conditioning on subsequent test levels of locomotor activity and conditioned place preference (CPP) was investigated by giving rats discrimination training with morphine and saline in the presence of different tactile floor cues in an open field. Groups of rats were trained under one of three levels of confinement (not confined, confined to 1/4, or confined to 1/16 of the open field) and tested for locomotor activity and CPP in the entire open field after receiving a saline injection. Confined rats subsequently spent less time on the morphine-paired floor cues (CS+) and were more active throughout the open field during the test than rats not confined during conditioning. Rats confined to the smallest area spent the least time on the CS+ and were the most active. It is suggested that confinement may preserve the novelty of the testing environment, which in turn may interfere with rats' usual responses to incentive drug-paired stimuli. These findings may have important implications for versions of the CPP technique in which training and testing environments differ considerably.