A novel version of the conditioned place preference (CPP) technique was used in an attempt to determine whether tactile stimuli previously associated with morphine elicit approach and sustained contact. Empirical support for this view has been equivocal, prompting some to question the validity of the CPP technique. In the present study, rats received, during conditioning, morphine (10 mg/kg, IP) paired exclusively with an open field floor made of four quadrants of one texture (CS+) and saline with another floor made of four quadrants of a different texture (CS-). On the test for CPP, rats were given saline and placed in an open field containing either 1, 2, or 4 quadrants of the CS+ (with 3, 2, 0 quadrants of the CS-, respectively). These animals showed high absolute CPP scores on the test, spending, on average, as much as 83% and 75% of their time on the CS+ when two and one CS+ quadrants, respectively, were present. Concurrent measures of activity indicated that animals were most active when all four quadrants were CS+ and least active when zero or one CS+ quadrant was present. Thus, once an animal approached and made contact with the CS+ it tended to maintain contact with this stimulus and to reduce its approach to and contact with other stimuli. The differentiating features of this version of the CPP technique, as well as the relationship between morphine-induced conditioned locomotion and CPP, are discussed.