The results of the Diabetes Intervention Study indicate that postprandial hyperglycaemia, but not fasting hyperglycaemia, is an independent risk factor for myocardial infarction and total mortality in newly detected non-insulin-dependent diabetes mellitus (NIDDM). Blood pressure and triglyceride levels were also found to be risk factors. Data from patients in the Study to Prevent NIDDM show that patients with impaired glucose tolerance (IGT) have higher fasting levels of triglycerides, and higher fasting and postprandial levels of pro-insulin, insulin, and C-peptide levels, compared with those who have normal glucose tolerance (NGT). Postprandial data at 2 h showed that hyperinsulinaemia was prolonged in subjects with IGT compared with their NGT counterparts. Postprandial pro-insulin levels at 2 h were also increased in subjects with IGT, and were found to have the best discriminating power between subjects with IGT and NGT. Those with IGT also secreted markedly higher levels of insulin, indicated by C-peptide levels, than those with NGT. These data support the possibility that subjects with IGT are at an increased risk of conditions such as atherosclerotic disease, even though they have not developed manifest NIDDM. Intervention with agents that can affect risk factors, such as triglyceride levels and postprandial hyperglycaemia, are expected to reduce the risk of atherosclerosis. One promising agent is acarbose, an antihyperglycaemic drug that significantly reduces postprandial insulin and triglyceride levels and has no effect on leptin levels. The ability of acarbose to improve glycaemic control without the potential for inducing hyperinsulinaemia and weight gain appears to give it an advantage over oral antidiabetic agents such as the sulphonylureas.