Currently, women with gestational diabetes mellitus have a repeat oral glucose tolerance test 6 weeks after delivery to identify those with persistent glucose intolerance. In this study, 298 women with gestational diabetes had a postnatal oral glucose tolerance test and of these, 23.2% had persistent glucose intolerance after delivery. The aim of this study was to determine if the antenatal oral glucose tolerance test results could be used to predict which patients would have persistent glucose intolerance after delivery. If only those with severely abnormal antenatal results had persistent disease, then those with only mildly abnormal antenatal results could be spared the postnatal oral glucose tolerance test. Using receiver operator characteristic curves, this study showed that the antenatal oral glucose tolerance test results could not be used to predict reliably which patients would continue to have abnormal glucose tolerance postnatally. Hence, postnatal oral glucose tolerance testing of all patients with gestational diabetes is still necessary as the detection of persistent glucose intolerance is important for the control of the diabetic condition so as to minimise long-term complications.