Background: The aim of this study is to investigate the prevalence of persistent postpartum glucose intolerance and to examine antepartum clinical characteri- stics for their predictability of persistent postpartum glucose intoleranc#e in the patients with GDM. Methods: In 211 GDM patients who showed more than two abnormal glucose values of O'Sullivan and Mahan's criteria on 100g-oral glucose tolerance test (OGTT), 75g-OGTT were performed at 6 weeks postpartum. The incidence of postpartum normal glucose tolerance (NGT), impaired glucose tolerance (IGT), and diabetes mellitus (DM) were investigated and antepartum ciinical parameters were compared among the three groups, Predictability of antepartum clinical characteristics for postpartum IGT and DM were also investigated by logistic regression analysis. Results: When we grouped the patients into postpartum NGT, IGT, DM according to the results of 75g-OGTT performed 6 weeks postpartum, The incidence were 81,5% of subjects had NGT, 9.0% had IGT, and 9.5% had DM. Plasma glucose levels and GAUC on antepartum 100 g-OGTT(NGT: 1660±159, IGT: 1948±730, DM: 2538±629mmol/L ' min), and proportion of patients receiv- ing insulin therapy increased progressively and significantly in association with worsening postpartum glucose tolerance. Frequency of positive family history of DM in qroups with IGT and DM (63,2% & 80.0%) were significantly higher than that in group with NGT(37,2%). Weight gain before diagnosis of GDM in groups with IGT and DM(6.7±3.9kg & 6.8±4.1 kg) were significantly smaller than that of group with NGT(9.5±3,5kg), Gestational age at diagnosis of GDM in group with DM(25.8±5.4 weeks) was significantly shorter than that in group with NGT(30.0±3,3 weeks), Proportion of subjects diagnosed earlier than 24 weeks of gestation were significantly higher in groups with IGT (15.8%) and DM (25.0%) than in group with NGT (1.2%). Proportions of subjects delivered heavier infants, ≥4 kg, were significantly higher in the DM group (40.0%) than in the NGT group (9.3%). In the patients having fasting plasma glucose levels hlgher than 5.8 mmol/L on antepartum 100g-OGTT, the prevalence of persistent glucose intolerance was significantly higher than in the patients FPG level lower than 5.8 mmol/L (61.9% vs 7.7%), Logistic regression analysis were performed using IGT and DM as the outcome of interest. The GAUC on antepartum 100g-OGTT, family history of DM, and the gestational age at diagnosis of GDM were independent predictors for both postpartum DM and postpartum IGT. Conclusion: The prevalence of persistent postpartum glucose intolerance in GDM patients were 18.5% and the most important independent predictor for persistent postpartum glucose intolerance was the degree of severity in glucose intolerance during pregnancy ( J Kor Diabetes Asso 24 : 46 - 59, 2000 ).