Cannabinoid CB1 receptors emerge early in the rat brain during prenatal development, supporting their potential participation in events related to neural development. In the present investigation, we completed earlier studies, analyzing CB1 receptor binding and mRNA expression by using autoradiography and in situ hybridization, respectively, in the brain of rat fetuses at gestational day (GD) 21 and of newborns at postnatal days (PND) 1 and 5, in comparison with the adult brain. These analyses were paralleled by quantitation of levels of anandamide and its precursor, N-arachidonoyl-phosphatidylethanolamine (NAPE), and of 2-arachidonoyl-glycerol (2-AG), carried out by using gas chromatography / mass spectrometry of the tri-methyl-sylyl-ether derivatives. As expected, CB1 receptor binding was detected at GD21 in a variety of brain structures. In most of them, such as the hippocampus, cerebral cortex, cerebellum, basal ganglia, and limbic nuclei, there were no marked differences in the density of CB1 receptors in animals at GD21 as compared to early newborns (PND1 and 5), although it markedly increased in these regions in adulthood. However, with the exception of the cerebellum and, in part, the caudate-putamen, the pattern observed for binding in these regions was clearly different from that observed for mRNA expression of the CB1 receptor, which currently exhibited the highest levels at PND1 and the lowest in the adult brain. This was also seen in the basolateral amygdaloid nucleus, ventromedial hypothalamic nucleus, medial habenula, and other structures. In the caudate-putamen and, particularly, in the cerebellum, mRNA expression was higher in the adult brain as compared with other ages. As previously reported, specific binding for CB1 receptors was also detected at GD21 in white matter areas, such as the corpus callosum, anterior commissure, fornix, fimbria, stria medullaris, stria terminalis, and fasciculus retroflexum. With the exception of the anterior commissure and the fimbria, specific binding progressively decreased at PND1 and PND5 until disappearing in the adult brain. In the fimbria, the highest values of binding were seen at PND1, but binding also completely disappeared in the adult brain, whereas in the anterior commissure, specific binding at PND1 and PND5 was lesser than that observed at GD21 and, particularly, in adulthood. CB1 receptor mRNA expression was not detected in these white matter areas, thus dismissing the possible presence of these receptors in glial cells rather than in neuronal axons. However, mRNA expression was detected in the brainstem, an area also rich in white matter, and it mostly correlated with receptor binding, exhibiting a progressive decrease from GD21 up to adulthood. CB1 receptor mRNA expression was also detected at GD21 in atypical areas where binding was not detected. These areas are proliferative regions, such as the subventricular zones of the neocortex, striatum, and nucleus accumbens. This atypical location only persisted at PND1 and PND5 in the striatal subventricular zone, but disappeared in the adult brain. We also found measurable levels of different endogenous cannabinoids in the developing brain. High levels of 2-AG, comparable to those found in the adult brain, were measured at GD21, whereas significantly lower levels were measured for anandamide and NAPE at this fetal age compared with the levels found in the adult brain. Levels of anandamide and NAPE increased during the early postnatal period until reaching the maximum in the adult brain. By contrast, 2-AG levels peaked at PND1, with values approximately twofold higher than those found at the other ages. In summary, all these data demonstrate that the endogenous cannabinoid system, constituted by endogenous ligands and receptor signaling pathways, is present in the developing brain, which suggests a possible specific role of this system in key processes of neural development. Synapse 33:181–191, 1999. © 1999 Wiley-Liss, Inc.