ObjectiveTo assess the usefulness of a breakfast test in determining which women with gestational diabetes do not need self-monitoring of blood glucose levels (home monitoring).Methodsa 1-hour post-standardized breakfast blood glucose below 7.8 mmol/L (140 mg/dL) was measured in 227 women and at or above 7.8 mmol/L in 115. Within each group, women were randomized to home monitoring with a meter or to clinic follow-up. Target glucose values were 5.3 mmol/L (95 mg/dL) fasting, 5.6 mmol/L (101 mg/dL) before meals, and 7.8 mmol/L (140 mg/dL) 1 hour postprandial. Up to these thresholds women on clinic follow-up were transferred to home monitoring. Insulin therapy was started on the same thresholds in women randomized or transferred to home monitoring. Large ofr gestational age (LGA) newborns represented the main outcome, with the transfer rate to home monitoring and need of insulin therapy the secondary ones.ResultsThe LGA delivery rate was not significantly different in the two follow-up groups in women with a breakfast result below 7.8 mmol/L (9.8 versus 4.3%) but was higher in the clinic follow-up among women with a breakfast result at or above 7.8 mmol/L (13.3% versus 30.9%;P< .05). Fewer women with a breakfast result below 7.8 mmol/L were transferred to home monitoring (2.6 versus 52.7%;P< .001) or started on insulin therapy (3.6 versus 25.2%;P< .001). The breakfast test cutoff of 7.8 mmol/L predicted insulin need with a sensitivity of 91.0% and a specificity of 72.0%.ConclusionA breakfast test is useful in identifying a low-risk population in which clinic follow-up may be used safely.