Epidemiological studies have found coffee consumption is associated with a lower risk for type 2 diabetes mellitus, but the underlying mechanisms remain unclear. Thus, the aim of this randomised, cross-over single-blind study was to investigate the effects of regular coffee, regular coffee with sugar and decaffeinated coffee consumption on glucose metabolism and incretin hormones. Seventeen healthy men participated in five trials each, during which they consumed coffee (decaffeinated, regular (containing caffeine) or regular with sugar) or water (with or without sugar). After 1 h of each intervention, they received an oral glucose tolerance test with one intravenous dose of [1-13C]glucose. The Oral Dose Intravenous Label Experiment was applied and glucose and insulin levels were interpreted using a stable isotope two-compartment minimal model. A mixed-model procedure (PROC MIXED), with subject as random effect and time as repeated measure, was used to compare the effects of the beverages on glucose metabolism and incretin parameters (glucose-dependent insulinotropic peptide (GIP)) and glucagon-like peptide-1 (GLP-1)). Insulin sensitivity was higher with decaffeinated coffee than with water (P<0·05). Regular coffee with sugar did not significantly affect glucose, insulin, C-peptide and incretin hormones, compared with water with sugar. Glucose, insulin, C-peptide, GLP-1 and GIP levels were not statistically different after regular and decaffeinated coffee compared with water. Our findings demonstrated that the consumption of decaffeinated coffee improves insulin sensitivity without changing incretin hormones levels. There was no short-term adverse effect on glucose homoeostasis, after an oral glucose challenge, attributable to the consumption of regular coffee with sugar.