OBJECTIVE To investigate how glucose levels at diagnosis of gestational diabetes (GDM) are associated with infant birth weight and long-term risk of manifest diabetes mellitus in the mother. METHODS In a case control study GDM pregnancies (n=2085) were compared with non-GDM pregnancies matched for day of delivery and obstetric unit (n=3792). GDM was defined as capillary blood glucose (cB-glucose) ≥9.0mmol/l (plasma glucose ≥10.0mmol/l) after a 75g oral glucose tolerance test (OGTT). The GDM cohort were followed up 8.5-13.5yrs after initial diagnosis with a questionnaire, answered by 1324 GDM women (65%). RESULTS GDM women had higher mean infant birth-weight compared with controls (3682g vs. 3541g, P<0.001). In multiple linear regression analysis, birth weight was positively correlated to fasting cB-glucose at GDM diagnosis (P<0.001), increased week of gestation (P<0.001) and BMI before pregnancy (P<0.003), while 2h OGTT cB-glucose values ≥9.0mmol/l were not related. Infants born to mothers with fasting cB-glucose ≤4.5mmol/l had no increased mean birth-weight or macrosomia (≥4500g) compared to controls. In the follow up 334/1324 women (25%) of the GDM women had developed diabetes, 215 type 2 diabetes, 46 type 1 diabetes and 72 unclassified diabetes. In logistic regression fasting cB-glucose and 2h OGTT cB-glucose at diagnosis of GDM as well as BMI >25 and origin outside Europe were risk factors for manifest diabetes. CONCLUSIONS Fasting blood glucose at diagnosis of GDM gives important information besides 2h OGTT glucose about pregnancy outcome and future risk for maternal diabetes.