Aims: To assess whether early pregnancy HbA1c can predict gestational diabetes mellitus (GDM) and adverse birth outcomes in Australian women. Methods: Prospective study of 466 women without diabetes, aged ≥16-years at first antenatal presentation. Recruitment was from 27 primary healthcare sites in rural and remote Australia from 9-January 2015 to 31-May 2018. HbA1c was measured with first antenatal investigations (<20-weeks gestation). Primary outcome measure was predictive value of HbA1c for GDM, by routine 75 g oral glucose tolerance test (OGTT; ≥24-weeks gestation), and for large-for-gestational-age (LGA) newborn. Results: Of 396 (129 Aboriginal) women with routine OGTT, 28.8% had GDM (24.0% Aboriginal). HbA1c ≥5.6% (≥38 mmol/mol) was highly predictive (71.4%, 95% CI; 47.8–88.7%) for GDM in Aboriginal women, and in the total cohort increased risk for LGA newborn (RR 2.04, 95% CI; 1.03–4.01, P = 0.040). There were clear differences between Aboriginal and non-Aboriginal women: 16.3% v 5.2% (P < 0.001) had elevated HbA1c whereas 12.4% v 29.6% (P < 0.001) developed hyperglycemia during pregnancy. Conclusions: Early pregnancy HbA1c ≥5.6% (≥38 mmol/mol) identifies Aboriginal women with apparent prediabetes and elevated risk of having an LGA newborn. Universal HbA1c at first antenatal presentation could facilitate earlier management of hyperglycemia and improved perinatal outcome in this high-risk population.