OBJECTIVE: Large preloads of protein and fat have been shown to lower glucose after a carbohydrate-rich meal in people with type 2 diabetes but add a considerable energy burden. Low calorie preloads [<5% of daily energy intake] have been tested in this study in people with prediabetes and with type 2 diabetes. RESEARCH DESIGN AND METHODS: This was an unblinded randomised crossover study with two placebo days and two active treatment days. Glucose was measured for 3 hours with fingerprick samples as well as continuous glucose monitoring [CGMS]. Twenty-four subjects with pre-diabetes or moderately controlled type 2 diabetes [fasting glucose < 10 and HbA1c < 8.5%] were recruited. The preload contained 17 g whey protein plus 3 g lactose and 5 g guar, and 1 g flavour material [including sucralose] dissolved in 150 ml cold water or 150 ml cold water with no additives. The breakfast test meal consisted of 2 slices of bread, margarine and jam [3 slices for men] with the test drink 15 minutes beforehand. RESULTS: Peak fingerprick glucose was reduced by 2.1 mmol/L at 45 min [p < 0.0001]. Average fingerprick glucose over 3 hours was reduced by 0.8 mmol/L [p = 0.0003]. There was no difference between those with diabetes or prediabetes or those on medication or not on medication. CONCLUSIONS: An 80 kcal whey protein/fibre preload can lower average glucose over 3 hours by 0.8 mmol/L. If used long term before at least two carbohydrate-rich meals/day this preload could lower HbA1c by up to 1%.