OBJECTIVE To determine how the omission of the third hour glucose measurement of the 100 g oral glucose tolerance test (GTT) affects the diagnosis of gestational diabetes mellitus (GDM). METHODS Retrospective chart review of 876 women delivered in a tertiary care hospital in Israel during a three-year period, who underwent a 100 g 3 hour oral GTT following an abnormal 50 g glucose screen. GDM was diagnosed according to the "criterion standard" accepted in the Fourth International Workshop Conference on GDM. The results of the 100 g 3 hour oral GTT were then retrospectively re-evaluated by omission of the third hour plasma glucose measurement from the "criterion standard". RESULTS GDM was diagnosed in 28.4% of the study. patients, while the omission of the third hour glucose measurement resulted in a 26.4% diagnosis of GDM. The perinatal data of the 18-omitted cases suggests that their exclusion from the GDM group would not have altered substantially the perinatal outcome of the study cohort. CONCLUSIONS A 100 g 2 hour oral GTT is a simple and economic alternative to the 100 g 3 hour oral GTT.