In type 2 diabetes the effects of short-term calorie restrictions on glycaemia are usually judged on fasting plasma glucose. As fasting duration rarely exceeds 2–3 h, we determined the effects of calorie restriction over different daytime periods using a continuous glucose monitoring system (CGMS) in noninsulin-using patients with type 2 diabetes.Fourteen poorly controlled (mean HbA1c = 9·3%) overweight or obese patients (mean body mass index = 30·1 ± 0·7 kg m−2) with type 2 diabetes were investigated twice with a CGMS, at baseline and at the end of a 18-day calorie-restricted diet (1490 kcal day−1). Areas under curves (AUCs) of 24-h glucose monitoring were measured and divided into postprandial and interprandial AUCs according to meal times.Areas under curves of 24-h glucose monitoring were significantly decreased after dieting:–37%,P = 0·0047. Both post and interprandial AUCs were significantly diminished (−26%,P = 0·0186 and−48%,P = 0·0037, respectively), but changes in interprandial AUCs were more marked than changes in postprandial AUCs (P = 0·0060). Nycthemeral peaks of glucose were observed at mid-morning times and were not significantly different before (242 ± 15 mg dL−1) and after dieting (227 ± 16 mg dL−1).Short-term calorie restriction in overweight or obese patients with type 2 diabetes resulted in dichotomous responses between interprandial and postprandial glycaemic excursions. The resistance of mid-morning glucose peaking to calorie restriction should result in additional dietary or pharmacological measures at breakfast.Eur J Clin Invest 2005; 35 (4): 259 –264 [ABSTRACT FROM AUTHOR]