Sevoflurane is an extensively used anesthetic for pediatric patients, however, numerous studies showed that sevoflurane (SEVO) may cause long-term neurodevelopmental toxicity. Dexmedetomidine (DEX) has been shown to be protective against SEVO-induced neurotoxicity, but the mechanism remains unclear. The effects and mechanisms of different DEX administration routes on SEVO-induced neurotoxicity and long-term cognitive defects were determined and further investigated the role of sex in these processes. Male and female Sprague Dawley (SD) rats at postnatal day 7 (PND7) received an intraperitoneal injection of DEX (10 μg/kg) before or after exposure to 2.5% SEVO for 6 hours, or before and after SEVO exposure. The respiratory and mortality rates of the pups were recorded during anesthesia. Neuroapoptosis was evaluated by TdT-mediated dUTP Nick-End labeling (TUNEL) staining. Immunohistochemistry and immunofluorescence were employed to detect the expression of caspase-3 in neuronal cells and neurons. The expression of GSK-3β and DISC1 were determined by Western blotting or RT-qPCR. Morris Water Maze (MWM) test was used to evaluate the learning and memory ability of rats until they were 3 weeks and 5 weeks old. Compared with the control group, exposure to 2.5% SEVO resulted in increased neuroapoptosis, and decreased the expression of DISC1 at levels of mRNA and protein and phosphorylated GSK-3β in the developing brain. SEVO exposure during critical neurodevelopmental periods could cause persistent cognitive defects in adolescent male and female rats, and inhibited DISC1 and phosphorylated GSK-3β protein expression. The neurotoxic impacts of SEVO were lessened by the administration of DEX (10 μg/kg) before or after exposure. Our findings suggest that DEX (10 μg/kg) mitigates the neurotoxic effects of SEVO on the developing rat brain as well as postnatal cognitive defects by regulating the DISC1/GSK-3β signaling.