A standard breakfast was evaluated as a screening test (BT) for gestational diabetes in a case-control study. The breakfast had a high carbohydrate content and was designed based on women's preferences in a similar sample. Women were selected with a standard 50 g, 1 hour glucose screening test (GT). Fifteen women had a positive GT and 15 had a negative test. A glucose tolerance test (GTT) and a standard breakfast (719 kcal) were performed to all women. Age, weight, height, tricipital and subscapular skinfold thicknesses were obtained by previously standardized personnel and glucose levels were determined by an automatic technique (oxidase glucose). Ten women had gestational diabetes. Sensitivity and specificity were calculated for the GT and several values of the BT, in order to identify the threshold and optimum time. Concordance of both screening tests with GTT was calculated through Kappa test. For the BT, a sensitivity of 0.90 and specificity of 1.00 with a glucose level > or = 140 mg/L (7.2 mmol/L) at 60 and 90 minutes were found. The 60 minutes timing were preferred due to a shorter waiting time to performed the test. Concordance of GTT with GT and BT was 0.53 and 0.92 respectively, both statistically significant. GT had high frequency of false negative values (6/15) that explain the low kappa value. Standard breakfast can be used as an alternative method for assessing carbohydrate tolerance in pregnancy with a better physiological response in these women.