Exclusion of simple sugars from the diabetic diet is not always followed by patients and may not even be as crucial as was hitherto thought. We tested three types of mixed breakfasts (400 kcal, 50 g HCO) including an isoglucidic amount either of white bread (30 g), honey (20 g) or sucrose (15 g), at the critical morning period i.e. for breakfast, in a group of 21 Type 2 (non-insulin-dependent) diabetic patients (6 well- and 15 badly controlled). Mean plasma glucose and insulin levels were comparable on the three occasions: respectively with bread, sucrose and honey, peak glucose values were 18 mmol/l, 17.7 mmol/l and 17.5 mmol/l in the uncontrolled group versus 13.9 mmol/l, 12.8 mmol/l and 12.7 mmol/l in the well-controlled group. Peak insulin values were 33.6 mU/1,34.0 mU/l and 36.3 mU/l (p greater than 0.05) in uncontrolled patients against 57.5 mU/l, 54.8 mU/l and 52.5 mU/l in well-controlled subjects (p greater than 0.05). The mean increment in peak plasma glucose values for the three breakfasts was as follows: 6.9 mmol/l, 6.3 mmol/l and 6.2 mmol/l for the uncontrolled group against 7.2 mmol/l, 5.9 mmol/l and 6.2 mmol/l in well-controlled subjects; the mean increment in peak plasma insulin levels was 21.8 mU/l,22.0 mU/l and 24.2 mU/l in the controlled group versus 38.2 mU/l, 32.0 mU/l and 34.7 mU/l in the well-controlled subjects, all values being non-significantly different (p greater than 0.05). We conclude that, in acute conditions, simple sugars have no additional hyperglycaemic effect over an isoglucidic amount of bread in well-and in badly controlled Type 2 diabetic patients, even at breakfast.