To analyze the effects of substituting the National Diabetes Data Group (NDDG) criteria with the International Association of Diabetes and Pregnancy Study Groups (IADPSG) or American Diabetes Association (ADA) criteria for the diagnosis of early-onset gestational diabetes mellitus (Early-GDM) or first trimester abnormal glucose tolerance (1t-AGT). A retrospective cohort study was conducted of 3200 women: 400 with Early-GDM, 800 with GDM, and 2000 with Non-GDM, according to the NDDG criteria. Rates of women with missed and new Early-GDM according to the IADPSG or ADA criteria were calculated. Multivariate logistic regression analysis was used to compare perinatal outcomes between groups. Using the IADPSG criteria, 61.6% of women with Early-GDM according to the NDDG were undiagnosed (Missed-Early-GDM group), and 25.9% of women with GDM and 15.7% of women with Non-GDM were diagnosed with Early-GDM (New-Early-GDM groups). Perinatal outcomes were worse in Missed-Early-GDM than in Non-GDM and better in New-Early-GDM groups than in the Early-GDM group. According to the ADA recommendations, only 11.8% of women with Early-GDM according to the NDDG criteria were diagnosed. Replacing the NDDG recommendations for the diagnosis of Early-GDM with the IADPSG or ADA criteria would mean depriving a large number of women with AGT and higher risk of adverse perinatal outcomes from early treatment and treating others with lower risk.