Aims/hypothesis Small protein or lipid preloads are able to improve glucose tolerance to a different extent and through different and poorly defined mechanisms. We aimed at quantifying the effect of a mixed protein and lipid preload and at evaluating the underlying mechanisms. Methods Volunteers with normal (NGT, n=12) or impaired (IGT, n=13) glucose tolerance and patients with type 2 diabetes (n= 10) underwent two OGTTs coupled to the double glucose tracer protocol, preceded by either 50 g of parmesan cheese, a boiled egg and 300 ml of water, or 500 ml of water. We measured plasma glucose, insulin, C-peptide, glucagon-like peptide-1 (GLP-1), glucose-dependent insulinotropic polypeptide (GIP), pancreatic polypeptide (PP), NEFA and glucose tracers, and calculated glucose fluxes, beta cell function variables, insulin sensitivity and clearance. Results After the nutrient preload, the OGTT-induced rise of plasma glucose was lower than after water alone in each study group. This reduction-more pronounced across classes of glucose tolerance (NGT -32\%, IGT -37\%, type 2 diabetes -49\%; p < 0.002)-was the result of different combinations of slower exogenous glucose rate of appearance, improved beta cell function and reduced insulin clearance, in this order of relevance, which were associated with an only mild stimulation of GIP and GLP-1. Conclusions/interpretation After a non-glucidic nutrient preload, glucose tolerance improved in proportion to the degree of its baseline deterioration through mechanisms that appear particularly effective in type 2 diabetes. Exploiting the physiological responses to nutrient ingestion might reveal, at least in the first stages of the diabetic disease, a potent tool to improve daily life glycaemic control.