Experiment 1 examined whether microinjections of morphine (1 microg in 0.5 microl over 1 min x 2 pairings) into 13 different CNS sites produced a conditioned place preference (CPP). Injections into the lateral ventricles (LV), ventral tegmental area (VTA), or periaqueductal gray (PAG) produced a CPP; injections 1 mm dorsal to the PAG or VTA, or into the caudate putamen, medial frontal cortex, hippocampus, lateral nucleus of the amygdala, lateral hypothalamus, pedunculopontine tegmental nucleus, posterior hypothalamus, ventral palladium, or nucleus accumbens septi (core or shell) did not. In Experiment 2, morphine 0.2 microg produced a CPP when injected into the VTA but not in the PAG, while 5.0 microg was effective in both sites. The CPP induced by systemic morphine (4 mg/kg x 1 pairing) was blocked by naloxone methiodide (NM) injected (2 nmol in 0.5 microl) into the VTA. PAG injections of 2 nmol reduced, and 5 nmol NM eliminated, the CPP. The results confirm that morphine injections into the VTA or the PAG are rewarding, that blockade of opioid receptors in either site disrupts a morphine-induced CPP, and that the VTA is more sensitive to both effects.