Glucagon-like peptide 1 (GLP-1) has been proposed to act as an incretin hormone due to its ability to enhance glucose-stimulated insulin secretion. Because GLP-1 also decelerates gastric emptying, it physiologically reduces rather than augments postprandial insulin secretory responses. Therefore, we aimed to antagonize the deceleration of gastric emptying by GLP-1 to study its effects on insulin secretion after a meal. Nine healthy male volunteers (age 25 +/- 4 years, BMI 25.0 +/- 4.9 kg/m2) were studied with an infusion of GLP-1 (0.8 pmol.kg(-1).min(-1) from -30 to 240 min) or placebo. On separate occasions, the prokinetic drugs metoclopramide (10 mg), domperidone (10 mg), cisapride (10 mg, all at -30 min per oral), or erythromycin (200 mg intravenously from -30 to -15 min) were administered in addition to GLP-1. A liquid test meal (50 g sucrose and 8% mixed amino acids in 400 ml) was administered at 0 min. Capillary and venous blood samples were drawn for the determination of glucose (glucose oxidase), insulin, C-peptide, GLP-1, glucagon, gastric inhibitory polypeptide (GIP), and pancreatic polypeptide (specific immunoassays). Gastric emptying was assessed by the phenol red dilution technique. Statistical analyses were performed using repeated-measures ANOVA and Duncan's post hoc test. GLP-1 significantly decelerated the velocity of gastric emptying (P < 0.001). This was completely counterbalanced by erythromycin, whereas the other prokinetic drugs used had no effect. Postprandial glucose concentrations were lowered by GLP-1 (P < 0.001 vs. placebo), but this effect was partially reversed by erythromycin (P < 0.05). Insulin secretory responses to the meal were lower during GLP-1 administration (P < 0.05 vs. placebo). However, when erythromycin was added to GLP-1, insulin concentrations were similar to those in placebo experiments. The suppression of meal-related increments in glucagon secretion by GLP-1 was reversed by erythromycin (P < 0.001). The time course of GIP secretion was delayed during GLP-1 administration (P < 0.05), but when erythromycin was added, the pattern was similar to placebo experiments. GLP-1 administration led to a reduction in pancreatic polypeptide plasma concentrations (P < 0.05). In contrast, pancreatic polypeptide levels were markedly increased by erythromycin (P < 0.001). Intravenous erythromycin counteracts the deceleration of gastric emptying caused by GLP-1, probably by interacting with the parasympathetic nervous system (pancreatic polypeptide responses). Despite augmented rises in insulin secretion, the glucose-lowering effect of GLP-1 is markedly reduced when the deceleration of gastric emptying is antagonized, illustrating the importance of this facet of the multiple antidiabetic actions of GLP-1.