The study was designed to identify those pregnant women who are diagnosed as having gestational diabetes by National Diabetes Data Group (NDDG) criteria, but normal glucose tolerance (NGT) or impaired glucose tolerance (IGT) by the World Health Organization (WHO) criteria, and to test whether treatment changed the perinatal outcome in those with NGT and IGT. The 216 women with an abnormal 100 g oral glucose tolerance test (OGTT) using NDDG criteria were subjected to a 75 g OGTT. Using the WHO criteria, 111 women (51%) had NGT, 98 (45%) had IGT and 7 (3%) had frank diabetes mellitus. Those with NGT and IGT were randomized into control and treatment groups. The perinatal outcome in these two groups was comparable whether the NGT and IGT groups were analysed together or separately except, that in those who were treated for IGT, smaller babies were born one week earlier than in the control group (3407 g vs 3110 g, P less than 0.01). This suggests that the WHO criteria can safely replace the 100 g OGTT with substantial savings in manpower, money and patients' time.