This study demonstrates that patients who have undergone coronary artery bypass graft (CABG) surgery have a high prevalence of abnormal glucose tolerance 3 months and 1 year later. Although only 6% were known to have diabetes mellitus (DM) preoperatively, a further 4% were classified DM at two oral glucose tolerance tests (OGTTs) over the subsequent year and a further 18% were classified as having impaired glucose tolerance (IGT) at 12 months. Reproducibility of the 120-minute plasma glucose level in the 75-g OGTT was estimated from a repeat test performed within 10 days. The coefficient of variation (CV) of 120-minute glucose was between 14% and 18%. The observed changes in class of glucose tolerance observed at OGTTs repeated 6 and 12 months after surgery differed from the predicted changes based purely on the estimated variability of 120-minute glucose measurement. There was evidence of regression to the mean for the IGT group. However, there was also evidence of deteriorating glucose tolerance in some subjects. Between 4% and 9% of those with IGT 3 months after CABG surgery developed DM by 12 months, and 26 (13%) of those with initially normal glucose tolerance (NGT) developed IGT. Insulin and glucose responses in the OGTT and estimates of insulin resistance and beta-cell function from fasting samples show that insulin resistance was the principal abnormality in IGT subjects, whereas in DM subjects, both insulin resistance and beta-cell dysfunction contributed. Analysis of preoperative patient characteristics showed that the presence of either a systolic blood pressure of 140 mm Hg or body mass index (BMI) of more than 25 kg/m2 identified 51% of the subjects who would at 1 year after surgery include all those who would be classified DM and 67% of those who would have IGT. Further analyses including insulin levels identified groups at particularly high risk of DM, but no combination of readily available preoperative measures identified all those destined to be classified IGT.