The Diabetes Control and Complications Trial (DCCT) in type 1 diabetes, published in 1993, demonstrated the beneficial effects of intensive insulin therapy on the onset and progression of diabetic retinopathy, nephropathy and neuropathy (1). Although many clinicians assumed that these results were also applicable to type 2 diabetic patients, this has not been demonstrated so far. In fact, a previous clinical trial in 1970, the University Group Diabetes Program (UGDP), compared five different treatments (diet and placebo, standard insulin doses according to body weight, variable insulin doses according to glycemia, and tolbutamide or phenformin) in type 2 diabetics (2). During the 5.5 years (range 3–8 years) that the 1000 patients were followed, the UGDP revealed no statistically significant differences in cardiovascular mortality or microvascular disease between the different treatment groups, except for the phenforminand tolbutamide-treated patients whose cardiovascular mortality was significantly higher compared with the placebo group. These results generated great controversy about the use of oral hypoglycemic agents, which was fueled by potential methodological deficiencies. For example, 46% of the patients included in the study already had signs of pre-existing cardiovascular disease and potentially due to non-homogeneous randomization, the pheneformine and tolbutamide group had a higher incidence of myocardial infarctions. Despite a reassessment by an international committee of the Biometric Society in 1976 the controversy over these results remains. Since then, in vitro studies have added to the fear that sulfonylureas may aggravate cardiovascular disease (3). In the study from Kumamoto, Japan in 1995 the effects of intensive and conventional insulin therapy on microvascular complications were compared in 100 lean (body mass index 2% difference in mean HbA1c levels (7% versus 9%) over the whole study period. Intensive glycemic control delayed the onset and the progression of diabetic retinopathy, nephropathy, and neuropathy. No major effect on macrovascular disease was seen. Despite the data showing a reduction in microvascular complications by insulin therapy, the effect of oral antidiabetic therapy on these complications remained to be elucidated. In the mid 1970s, the prospective and randomized United Kingdom Prospective Diabetes Study (UKPDS) was designed to try to clarify whether intensive therapy as compared with dietary therapy had a better longterm outcome in type 2 diabetes. An additional aim was to compare the advantages and disadvantages of insulin and sulfonylureas among patients in the intensive therapy group. In obese patients, an additional group was assigned to metformin. Later on, a hypertension study was added to examine the effects of tight blood pressure control on the risk of macroand microvascular complications. Three aggregate endpoints were used to assess differences between the conventional and intensive treatments: (i) diabetes-related endpoints (sudden death, death from hyperglycemia or hypoglycemia, myocardial infarction, angina, heart failure, stroke, renal failure, amputation, vitreous hemorrhage, retinopathy requiring photocoagulation, blindness in one eye, or cataract extraction); (ii) diabetes-related death (death from myocardial infarction, stroke, peripheral vascular disease, renal disease, hyperglycemia or hypoglycemia, and sudden death); and (iii) all-cause mortality. The most important results of the UKPDS have now been published in four different papers in September 1998 (5–8). Before examining the results of the UKPDS, it should be noted that, unavoidably, considerable therapeutic overlap among the different treatment groups occurred (9). If in the control or intensive therapy groups fasting plasma glucose (FPG) levels were above predefined limits (FPG>15 mmol/l or FPG>6 mmol/l respectively), a sulfonylurea was added. This therapy could be further supplemented with metformin, and eventually replaced by insulin if necessary. Thus, in each treatment group at least 30% of patients took at least one additional drug than they were initially assigned to. In the UKPDS 33 publication (5), intensive therapy with a sulfonylurea (chlorpropamide, glibenclamide, or glipizide) or with insulin aimed at obtaining a FPG of less than 6 mmol/l was compared with conventional therapy (diet only with drugs added only if hyperglycemia was symptomatic or if FPG>15 mmol/l) by randomizing patients with a FPG>6 mmol/l to either treatment group. On recruitment 93% of the patients had a FPG>7.0 mmol/l (novel American Diabetes Association (ADA) criteria for the diagnosis of diabetes) H IG H L IG H T European Journal of Endocrinology (1999) 140 4–6 ISSN 0804-4643