Objective Determine the association of Non-alcoholic fatty liver disease(NAFLD) to the occurrence of GDM and miscarriages. Design Population-based prospective cohort study. Setting Anuradhapura district-Sri Lanka. Sample All pregnant women(gestational age<12 weeks) registered in the national pregnancy care programme(>98%coverage) during July to September-2019, recruited in the Rajarata Pregnancy Cohort(RaPCo)(included 80% of registered mothers). Method Clinical, biochemical parameters and ultrasound scan(USS) of liver were performed at the baseline to assess NAFLD. Excluding pregnant women who reported any type of hyperglycemia at the baseline, the cohort was followed up to assess the occurrence of GDM (using latest WHO criteria) during 24-28 weeks of gestation. Miscarriages were documented. Main outcome measures GDM and miscarriages. Results The prevalence of Fatty liver grade(FLG)-II and I was 14.2%(n=90), 37%(n=234), respectively. The incidence of GDM among the normoglycemic pregnant women with FLG-II, I, and 0 were 162.2, 43.9, 11 per 1000, respectively. After adjusting for age, BMI, and other known risk factors, women with FLG-II had a relative risk(RR) of 12.5(95%CI 2.2-66.4) for developing GDM compared to those with FLG-0. Pregnant women with FLG-I-(RR= 5.1,95%CI-1.7-15.1) and II-(RR=8.4,95%CI-2.6-27.1) had a very high risk of early pregnancy miscarriage compared to FLG-0. Conclusion FLG-2 is a significant risk factor for GDM and miscarriages. Incorporating pre-conceptional or early pregnancy identification of NAFLD using simple USS into routine care provision will enable early risk identification and appropriate action. Funding Accelerating Higher Education Expansion and Development(AHEAD) grant(World Bank-funded project through University grant commission-Sri Lanka) funded this study. Keywords–NAFLD, GDM, Pregnancy, Ultrasound-scan, Miscarriages