BACKGROUND: The prevalence of type 2 diabetes mellitus and of gestational diabetes mellitus is high in the Mexican population; thus, strategies to improve its detection and prevent obstetric and perinatal complications are essential. PATIENTS AND METHODS: During the period 2000-2004 a total of 8,074 pregnant women were studied from the day of performance, 50-g, 1 hour glucose screening test for gestational diabetes mellitus (ST-GDM) until the end of pregnancy using the same protocol: ST-GDM was performed immediately after 14 weeks of gestation in high-risk women and between 24-28 weeks in those with regular risk. Two weeks later 100-g, 3 hour oral glucose tolerance test (3h-OGTT) was performed in women with ST-GDM [symbol: see text] 130 but < 180 mg/dL, and patients were classified according to Freinkel's criteria in class A1, A2 and B1 gestational diabetes mellitus. All women received a diet according to their ideal weight for gestational age and, when necessary, insulin was added to achieve an adequate glycemic control. RESULTS: From 8,074 glucose screening tests for gestational diabetes mellitus, 37.2% (n = 2,997) were positive and 17.2% (n = 514) diagnostic for gestational diabetes mellitus (serum glucose [symbol: see text] 180 mg/dL). In 2,483 patients, ST-GDM was positive but negative for gestational diabetes (serum glucose >130 but <180 mg/dL); in 1,070 of them (43.0%) gestational diabetes mellitus was diagnosed by means of 3h-OGTT. The diagnosis of class A1 gestational diabetes mellitus was done in 908 patients (84.9%), A2 in 188 (11.8%) and B1 in 51 (3.2%). Patients with gestational diabetes diagnosed only by an abnormal ST-GDM were classified as class A1 gestational diabetes mellitus. Gestational age at diagnosis by ST-GDM was 12.4 +/- 4.7 weeks and 27.4 +/- 5.8 weeks when diagnosed by a 3h-OGTT. The total daily dose of insulin required was related to Freinkel's classification: 16.0% (n = 215/1,345) with class A1 gestational diabetes mellitus required 0.48 +/- 0.33 U/day/kg and 85.0% (n = 43/51) with class B1 needed 1.0 +/- 0.32 U/day/kg at the end of pregnancy, which occurred at 38.2 +/- 2.0 weeks in all 3 groups. In decreasing order of occurrence perinatal complications were: pregnancy-induced hypertension (specially in class B1 gestational diabetes mellitus), urinary tract infections and premature rupture of membranes. Birth weights either < 2,500 g or > 4,000 g were more frequent among women with class B1 gestational diabetes, than in the other 2 groups. After pregnancy ended (6-8 weeks), 52.2% of women with gestational diabetes (mostly class B1) was diagnosed with permanent type 2 diabetes mellitus. DISCUSSION: In this study the percentage of pregnant women diagnosed as having gestational diabetes mellitus exclusively by ST-GDM was high (17.2%). Freinkel's classification of such disease predicted higher risk in women with class B1 gestational diabetes mellitus for: 1) obstetric complications, 2) higher doses of insulin to achieve an adequate glycemic control; 3) newborns' birth weight < 2,500 g or > 4,000 g for gestational age, 4) having type 2 diabetes mellitus in late puerperium.