Objectives: Gestational diabetes is a disease of two generations, affecting both the pregnant women and her babies. Gestational diabetes increases the risk of perinatal adverse outcomes and the incidence of diabetes after delivery. The diagnosis of gestational diabetes has revised several times. According to the updated criteria, the prevalence of gestational diabetes is high (17.8%). The diagnostic cutoffs are based on the studies with most of the study subjects to be Caucasians. However, our preliminary results from a retrospective study showed different relationships between plasma glucose and perinatal outcomes in women in Taiwan, compared with the findings from Caucasians. Therefore, the specific aim 1 in this study is to establish a cohort for gestational diabetes, in order to find cutoffs suitable for Asians. Besides, oral glucose tolerance tests are used in the diagnosis of gestational diabetes. However, it is inconvenient and costly. In non-pregnant adults, we have shown that hemoglobin A1c can reduce the need of oral glucose tolerance tests. Nonetheless, it is not useful in pregnant women. We have established an assay to measure serum glycated albumin, which stands for the average plasma glucose in the past 2-4 weeks. The specific aim 2 in this study, is to evaluate if glycated albumin can be used to diagnose gestational diabetes, and if glycated albumin can reduce the need of oral glucose tolerance tests. Moreover, we will measure blood metabolome in these pregnant women by global profiling metabolomic methods, which includes 380 metabolites. We will search for metabolites which can predict or diagnose gestational diabetes. The findings will lead to further studies on the pathogenesis of these metabolites. Besides, we will measure metabolites ever reported in the literature using targeted metabolomics approach. Their performance to diagnose and predict gestational diabetes will be evaluate. Research design and methods: This project will be completed in 3 years. In the first two years, we will recruit 1000 pregnant women to establish a cohort of gestational diabetes. Gestational diabetes will be diagnosed according to the recommendations by the IADPSG. Fasting plasma glucose will be measured at first visit. At 24-28 weeks, oral glucose tolerance tests will be done. Besides, we will record all the perinatal outcomes and potential confounders for gestational diabetes and perinatal adverse outcomes. In the second and third year, we will measure serum glycated albumin and metabolome, followed by statistic analyses. Glycated albumin will be measured by an enzymatic assay. A standard curve will be created to estimate the concentrations and percentages of glycated albumin in the samples. Metabolome will be measured by LC-QTOP MS and GC-TOP MS. The relationship between plasma glucose and perinatal adverse outcomes will be analyzed by scatter plots and logistic regression analyses adjusting for confounders. A detailed search will be performed to analyze the effect of different cutoffs on the prediction ability of perinatal outcomes, in order to develop cutoffs suitable for Asians. The relationships among glycated albumin, metabolome, and perinatal adverse outcomes will be analyzed by logistic regression, adjusting for confounders. A search for different cutoffs will be done to evaluate if glycated albumin and metabolome can be used to diagnose or improve the diagnosis of gestational diabetes.