1. In order to assess the efficacy of the use of the diurnal plasma glucose profile rather than that of the glucose tolerance test (GTT) to predict hyperglycemia during pregnancy, we compared the results of the two tests. A total of 192 pregnant women seen at the Prenatal Clinic of the Faculty of Medicine of Botucatu were submitted to the glucose tolerance test (GTT) and determination of diurnal plasma glucose profile. 2. On the basis of two blood tests (GTT and diurnal plasma glucose profile), the subjects were divided into four groups: Group I-A, normal GTT and profile (79 patients, 41.2%); Group I-B, normal GTT and altered profile (63 patients, 32.8%); Group II-A, altered GTT and normal profile (18 patients, 9.4%); Group II-B, altered GTT and profile (32 patients, 16.7%). 3. Large babies were delivered by 25.6% of Group I-A, 53.8% of Group I-B, 28.6% of Group II-A and 51.9% of Group II-B patients. Group I-A patients are normoglycemic, Group I-B patients have intolerance to carbohydrates, protein and lipids, Group II-A patients have intolerance to high carbohydrate amounts, especially in the form of glucose, and Group II-B patients are diabetic. 4. We propose that Group I-A patients should receive no treatment, Group II-A patients should be advised to avoid excess carbohydrate intake and Groups I-B and II-B patients should be placed on a low-calorie diet and treated with insulin if necessary to obtain normal blood glucose levels. 5. Routine determination of blood glucose levels under fasting conditions represents a screening method for diabetes and values of greater than or equal to 90 mg/dl identify a population at risk that should be submitted to GTT and determination of plasma glucose profile.