Background: The genetic basis for hyperglycaemia in pregnancy remain unclear. This study aimed to uncover the genetic determinants of gestational diabetes mellitus (GDM) and investigate their applications.Methods: We performed a meta-analysis of genome-wide association studies (GWAS) for GDM in Chinese women (464 cases and 1,217 controls), followed by de novo replications in an independent Chinese cohort (564 cases and 572 controls) and <i>in silico</i> replication in European (12,332 cases and 131,109 controls) and multi-ethnic populations (5,485 cases and 347,856 controls). A polygenic risk score (PRS) was derived based on the identified variants.Results: Using the genome-wide scan and candidate gene approaches, we identified four susceptibility loci for GDM. These included three previously reported loci for GDM and type 2 diabetes mellitus (T2DM) at <i>MTNR1B</i> (rs7945617, odds ratio [OR], 1.64; 95% confidence interval [CI], 1.38 to 1.96), <i>CDKAL1</i> (rs7754840, OR, 1.33; 95% CI, 1.13 to 1.58), and <i>INS-IGF2-KCNQ1</i> (rs2237897, OR, 1.48; 95% CI, 1.23 to 1.79), as well as a novel genome-wide significant locus near <i>TBR1-SLC4A10</i> (rs117781972, OR, 2.05; 95% CI, 1.61 to 2.62; <i>P<sub>meta</sub></i>=7.6×10<sup>-9</sup>), which has not been previously reported in GWAS for T2DM or glycaemic traits. Moreover, we found that women with a high PRS (top quintile) had over threefold (95% CI, 2.30 to 4.09; <i>P<sub>meta</sub></i>=3.1×10<sup>-14</sup>) and 71% (95% CI, 1.08 to 2.71; <i>P</i>=0.0220) higher risk for GDM and abnormal glucose tolerance post-pregnancy, respectively, compared to other individuals.Conclusion: Our results indicate that the genetic architecture of glucose metabolism exhibits both similarities and differences between the pregnant and non-pregnant states. Integrating genetic information can facilitate identification of pregnant women at a higher risk of developing GDM or later diabetes.