To assess long-term metabolic consequences of recurrent ictal events arising during development, seizures were repeatedly generated in rats at different stages of cerebral maturation. Seizures were induced by i.p. injections of bicuculline for three consecutive days, starting from postnatal day 5 (P5), when the brain is very immature, or from P15, a period at which the brain is more structurally organized. Local cerebral metabolic rates for glucose were measured in 74 structures at P15, P25 and in adults (P60), by the autoradiographic method using 2-D-[14C]deoxyglucose. Repeated seizures in P5 to P7 pups led to a reduction (16-34%) of glucose consumption at P15, mainly significant in sensory, motor and functionally non-specific areas as well as in cerebellar nuclei. Selective decreases in metabolic activity were still recorded in adults, mostly in auditory system (20%) and cerebellar nuclei (27%). Seizures generated from P15 to P17 led to an overall mortality rate of 62% (versus 22% at P5 to P7). Surviving animals exhibited reduced metabolic rates for glucose (by 7-27%) at P25, significant in 23 structures, and depicting pronounced changes in limbic, hypothalamic, sensory and white matter areas, whereas brain functional activity finally returned to basal values at P60. Therefore, while younger rats seemed to better tolerate repeated bicuculline-induced seizures than older animals, the reverse was true for long-term metabolic effects, and the more immature the brain when seizures arise, the more persistent the functional consequences.