The short- and long-term consequences of a neonatal exposure to diazepam (DZP) on the postnatal changes in local cerebral metabolic rates for glucose (LCMRglcs) were studied by the quantitative autoradiographic [14C]2-deoxyglucose method in a total number of 66 brain structures of freely moving rats. Rat pups received a daily subcutaneous injection of 10 mg/kg DZP, of the dissolution vehicle or of saline from postnatal day 2 (P2) to 21 (P21). The animals were studied at 4 ages, P10, P14, P21 and P60. DZP induced a decrease in LCMRglcs which was restricted to 13 areas at P10, mainly sensory and limbic regions. At P14, the treatment had significant metabolic effects on 48 structures belonging to all functional systems. By P21, 23 brain areas were still affected by the treatment, mainly sensory, limbic and motor areas. At P60, i.e. at about 40 days after the end of drug exposure, LCMRglcs still decreased in 14 brain regions which were mainly sensory and limbic structures. The structures most sensitive to both short- and long-term consequences of the anticonvulsant treatment are mammillary body, limbic cortices and sensory regions. The dissolution vehicle increased LCMRglcs in a few brain regions at P14 and P60, whereas it decreased metabolic levels in 5 brain regions at P21. The results of the present study show that the brain appears to be particularly vulnerable to the treatment at P14, period of active brain growth, whereas by P21, the drug is actively metabolized and a tolerance to the treatment may occur. The long-term effects of the treatment are in good accordance with the well-known effects of DZP on anxiety, sedation and memory. The structures most sensitive to early neonatal DZP exposure are the mammillary body, limbic cortices and sensory regions that all contain a high density of benzodiazepine binding sites.