Objective To investigate whether the long-term cognitive function will be affected by neonatal repeated anesthesia with ketamine in mice, and to explore the potential role of microglia and CX3CR1 signaling pathway in the process.Methods The postnatal day 6 (P6) littermates were randomly divided into the control group and the ketamine group. The ketamine group received intraperitoneal injection of 80 mg/kg ketamine once a day for five consecutive days, whereas the control animals were injected with an equal volume of saline. After one month and two months of administration, the Morris water maze test was performed respectively to reflect the spatial learning and memory abilities of the school-age and adult mice. The alterations in the number of microglia in the dentate gyrus of hippocampus were observed and the expression of CX3CR1 protein in bilateral hippocampal tissues was detected.Results The escape latency as detected in the Morris water maze test was significantly different among the distinct time points (1st, 2nd, 3rd, 4th and 5th day before training) in school-age and adult mice (F =34.213 and 41.527, both P < 0.001). The escape latency of school-age and adult mice was different between the ketamine group and the control group (F = 19.856 and 21.365, both P < 0.001), and was even greater in the ketamine group (P < 0.05). In addition, the change trend of the escape latency of school-age and adult mice was also different between the ketamine group and the control group (F = 12.436 and 17.548, both P < 0.001). The number of platform crossing within 60 s was greater, and the time spent in the target quadrant was longer in the control group relative to the ketamine group (P < 0.05). The number of microglia and the level of CX3CR1 protein in hippocampal dentate gyrus at P11, P16, P21, P35 and P67 were higher in the control groups than those in the ketamine group (P < 0.05).Conclusions Neonatal repeated anesthesia with ketamine leads to long-term cognitive dysfunction in mice, which may be associated with the decrease in the number of microglia and the expression of CX3CR1 protein in the hippocampus.