background: Recently, both trivalent chromium Cr (III) and hexavalent chromium Cr (VI) have been reported to produce neurotoxicity. However, the underlying mechanisms of the neurotoxicity caused by different chemical valence of chromium remain unclear. objective: The purpose of this study was to investigate the mechanism of neurotoxicity induced by exposure to chromium with different valence states based on metabolic disturbance in zebrafish larvae. methods: Zebrafish embryos were exposed to 1 mg/L Cr (III) and 1 mg/L Cr (VI) for 120 hpf respectively. The related indexes of neural development were observed by stereoscope and behavior analysis system. 8OH-dG were detected using enzyme-linked immunosorbent assay. The generation of reactive oxygen species was detected using an oxidant-sensing probe 2',7'-dichlorodihydrofluorescein diacetate. AChE activity was determined by a colorimetric assay based on hydrolysis of acetylcholine. The expression levels of neurodevelopmental genes and methyltransferase genes in juvenile zebrafish was analyzed by real-time PCR. The methylation status of neurogenin1 and neurod1 genes was detected by bisulfite sequencing PCR. The binding of H3K27me3 was detected by chromatin immunoprecipitation-qPCR. Metabolic profiles and one carbon metabolic analysis were performed by UPLC-MS. results: There were no significant differences in survival rate, hatching rate and spontaneous movement of zebrafish in both Cr-exposed groups compared to the control. The malformation rate in Cr (VI) -exposed group was obviously increased compared to the control and Cr (III) -exposed group. At 48hpf and 72hpf of exposure, the embryonic heart rate in Cr (III)-exposed group was significantly higher than that of Cr (VI)-exposed group and the control. At 120hpf, zebrafish in both Cr-exposed groups exhibited decreasing changes in swimming distance and disturbance of sensitivity to light and dark. 8OH-dG in Cr (VI)-exposed group were significantly higher than that in the control. The generation of ROS in both Cr -exposed groups was significantly higher than that in the control. The activity of AchE was significantly decreased in both Cr-exposed groups compared to the control. Most of early neurogenesis related genes, such as α-tubulin, elavl3, gap43, sox19b, neurogenin1 and neurod1 in Cr-exposed groups were significantly up-regulated compared to those in the control. The expression of dnmt1 and dnmt3 genes was significantly down-regulated in both Cr-exposed groups. BSP-PCR results showed that genic sequences in the neurogenin1 and neurod1 genes have lower levels of DNA methylation in both Cr-exposed groups, especial in Cr (VI)-exposed group. ChIP analysis showed that there was a decrease in H3K27me3 binding within the corresponding region of neurogenin1 in both Cr-exposed groups and that of neurod1 in Cr (III)-exposed group. Untargeted metabolomic analysis showed that significant changes in metabolites induced by Cr exposure were associated with differences in primary bile acid biosynthesis, phospholipid biosynthesis (phosphatidylcholine biosynthesis and phosphatidylethanolamine biosynthesis), linoleic acid metabolism, arachidonic acid metabolism, amino acid metabolism, purine metabolism, betaine metabolism, spermidine and spermine biosynthesis, and folate metabolism, the last four of which are related to one carbon metabolism. Targeted analysis of one carbon metabolites (5-MT, Gly, Met, SAH and Hcy) related with folate cycle and methionine metabolism were significantly decreased upon Cr exposure. The elevated SAM to SAH ratio in both Cr- exposed group indicated the decreasing capacity for methylation reaction. conclusions: Cr (III) and Cr (VI) can induce neurotoxicity by interfering with one carbon metabolism and affecting DNA methylation and histone methylation to regulate the expression of neuro-related genes. Cr exposure also influenced primary bile acid biosynthesis and phospholipid biosynthesis, which are associated with neuroprotective effects and need to be further validated.