Abstract: The nearly complete mitochondrial genome of Eucryptorrhynchus scrobiculatus was sequenced by next-generation sequencing (NGS) method, and constructed the phylogenetic trees using Maximum Likelihood and Bayesian inference methods, based on the nucleotide sequences from mitochondrial genome of 37 genes and the amino acid sequences from thirteen protein-coding genes of 29 Curculionidae species (ingroup) and 3 Chrysomelidae species (outgroup). The findings revealed that most of protein-coding genes start with the codons of ATN except for the nad1 genes. The latter one protein-coding genes use TTG as start codons. While cox1 and nad4 employ the incomplete T as the stop codon, the remaining nine protein-coding genes terminate with TAA or TAG. All tRNA genes can be folded into typical cloverleaf structure, with the except of the trnS1-AGN. Both phylogenetic inference methods produced a similar tree topological structure: the subfamilie Cryptorhynchinae, Ceutorhynchina, Molytinae, Scolytinae, Entiminae and Dryophthorinae were monophyletic. However the subfamilies Curculioninae were non-monophyletic. The E. scrobiculatus is sister to Trigonopterus carinirostris + (T. jasminae + T. kotamobagensis). In this paper, we obtained the complete mitochondrial genome sequence of the E. scrobiculatus, reconstructed the phylogeny of the Curculionidae family, and clarified the phylogenetic position of the E. scrobiculatus, which is beneficial to a more comprehensive understanding of the phylogeny and mitochondrial genomics of the Curculionidae family.