This clinical trial aims to test whether glibenclamide is non-inferior to insulin for the treatment of gestational diabetes.

In patients with gestational diabetes, a therapeutic treatment with a diet, a blood glucose monitoring and, if necessary, treatment with insulin is associated with reduced neonatal complications. Although the effectiveness of insulin is proven, this therapy has many drawbacks, especially in terms of implementation and monitoring by women. An alternative is the use of oral antidiabetic agents like glibenclamide.

Assumption: Glibenclamide is non-inferior to insulin for the treatment of gestational diabetes.

Main objective: To test the oral Glibenclamide is non-inferior to subcutaneous insulin for the occurrence of perinatal complications in the management of pregnant women developing gestational diabetes and requiring treatment other than dietary.

Secondary Objective: To test the noninferiority of two treatments on maternal glycemic control and maternal complications.

Sample size: With a 80% power and a significance level of 5%, 450 subjects per group are required to show that treatment with glibenclamide is not considered inferior to treatment with insulin if the frequency of outcome composite does not exceed 25% with glibenclamide, whereas it was 18% with insulin (calculated with a non-inferiority margin of 7% and considering that 20% of patients treated with glyburide and Insulin for change half of the patients accept randomization).

Progress of the study: Inclusion of women between 24 and 34 weeks. Randomization between Insulin and Glibenclamide after failure of 10 days of dietary treatment well done. Failure defined by at least 2 abnormal values of glycemia: fasting glucose ≥ 0.95 g / l and/or a postprandial 2h ≥ 1.20 g / l.

Women will receive either insulin according to the usual protocol or Glibenclamide at an initial dose of 2.5 mg / once daily in the morning before breakfast. So long as the glycemic targets will not be reached doses are gradually increased stepwise every 4 days until day 21 (20mg/j) . Patients who have not reached the glycemic targets at this highest dose will have a change in treatment for insulin at day 21. Routine monitoring of pregnant women with gestational diabetes, up to a week after delivery Number of clinical sites: 15 Perspective: Use of Glibenclamide as a first line treatment of gestational diabetes requiring drug treatment.

Inclusion Criteria:

• gestational diabetes diagnosed between 24 and 34 weeks according to international criteria.

Exclusion Criteria:

• Multiple Pregnancy
• Chronic Hypertension
• Preeclampsia
• Renal impairment
• Hepatic insufficiency
• Long time corticosteroids treatment
• Allergy to sulfa drugs
• Pre-existing diabetes in pregnancy
• Abnormal result on screening test for gestational diabetes before 24SA
• Fasting glucose ≥ 1.26 g / l at initial diagnosis of diabetes
• The need for drug treatment contraindicated or not recommended with taking Glibenclamide
• Poor understanding of French
• Lack of Social Insurance