In animal studies, the whole-body glucose disposal effect of insulin is low in the fasted state or after atropine infusion, but doubles after a meal, consistent with the hepatic insulin-sensitizing substance (HISS) hypothesis. We tested how a standardized test meal and atropine affected the dynamic response to insulin in humans. Insulin sensitivity was assessed in healthy male subjects (aged 28.9 +/- 1.9 years, body mass index 23.3 +/- 0.8 kg.m(-2)) by using the rapid insulin sensitivity test (RIST), which is a transient euglycemic clamp. After a 24-hour fasting period, dynamic insulin sensitivity was assessed and then repeated 100 min after the test meal. In a second protocol, the volunteers were fed the standardized test meal and intravenous atropine (0.5 mg) or saline (control group) was administered 50 min before insulin sensitivity assessment. Insulin sensitivity increased in the fed state (232.1% +/- 46.3%, n = 7) in comparison with the 24-hour fasted state. In the atropine protocol, the drug partially blocked (56.5% +/- 11.6%, n = 6) insulin sensitivity. In humans, feeding resulted in increased insulin sensitivity. The low dose of atropine in humans lead to a partial HISS-dependent decrease in insulin sensitivity. Meal- induced insulin sensitization occured in humans by a similar mechanism as that reported in other species. The sensitization process was regulated by a cholinergic 'feeding signal.'