Abstract:[Background] The oral microbiota changes have significant correlations with oral diseases and systemic diseases in pregnant women. [Objective] To compare the oral microbiota structure of females in different pregnancy states and explore the correlation between the structure difference and different pregnancy states. [Methods] Eighteen pregnant women and nine non-pregnant women were selected for the study and assigned into three groups (n=9):gestational diabetes mellitus (GDM, mean age of (28.9±3.6) years, mean pregnancy of (30.1±3.2) weeks), non-diabetic pregnant women (PW, mean age of (27.9±3.0) years, mean pregnancy of (28.6±4.7) weeks) and non-pregnant women (NPW, mean age of (27.7±2.1) years). Oral saliva (S) and supragingival dental plaque (D) samples were collected. The Illumina Novaseq platform was used to sequence the V3-V4 variable region of bacterial 16S rRNA. SILVA was used for the taxonomical annotation of the characteristic sequences, and QIIME was employed to perform the bioinformatics analysis. [Results] The D-GDM samples had higher detection rate of Capnocytophage and lower detection rate of Selenomonas than the D-PW samples. The S-GDM samples had higher detection rate of anaerobic bacteria such as Aeromonas and Bacteroides than the S-PW samples and lower detection rate of Veillonella than the S-PW samples. The D-PW samples had higher detection rates of Leptotrichia, Prevotella, and Selenomonas and lower detection rates of Aeromonas, Actinobacillus, Capnocytophage, Neisseria, Lautropia, and Streptococcus than the D-NPW samples. The S-PW samples had lower detection rate of Veillonella, Prevotella, Streptococcus, and Porphyromonas than the S-NPW samples. The D-GDM samples had high detection rate of Tannerella and Leptotrichia and lower detection rate of Roseburia than the D-NPW samples. The S-GDM samples had high detection rates of Treponema and Campylobacter and lower detection rate of Streptococcus than the S-NPW samples. At the species level, the S-GDM samples had higher detection rate of Fusobacterium nucleatum, similar detection rates of Porphyromonas gingivalis and Tannerella forsythia, and lower detection rate of Prevotella nigrescens than the S-PW samples. The S-PW samples had high detection rates of P. gingivalis, P. nigrescens, and T. forsythia than the S-NPW samples.[Conclusion] GDM will increase the detection rate of anaerobic bacteria in the oral cavity, and the correlations between GDM and periodontitis-causing bacteria such as P. gingivalis, T. forsythia, and P. nigrescens are uncertain.