We investigated the efficacy of intrapulmonary administration of short-acting porcine insulin in anaesthetized pigs (n = 14) in a randomized intervention study. Insulin was administered by a new jet nebulizer (Maxin) in a random order at different doses, 0 (saline), 10 or 40 U. The hypoglycaemic effect was compared to control (0.9% saline). Blood glucose and serum insulin concentrations were followed at specified time intervals for 90 min. Plasma catecholamine concentrations were measured in order to estimate the concurrent stress. Nebulized insulin caused a significant decrease in blood glucose concentrations (p < 0.0001) (n = 28) at all doses used. The decrease in mean blood glucose concentration from the start of nebulization was 39 +/- 3% (mean +/- SEM), falling from 4.6 +/- 0.1 to 2.8 +/- 0.2 mmol 1(-1), with a nadir at 40 min after the 40 U insulin dose (n = 10). Serum insulin concentration rose from (mean +/- SEM) 5.2 +/- 0.1 to 25 +/- 9 mU 1(-1) after the insulin dose of 40 U (n = 10), the peak value occurred at 30 min. The plasma catecholamine concentrations increased significantly (p < 0.0001) (n = 28) from 0 to 60 min, this increase was similar for control and for different insulin doses. We conclude that intrapulmonary administration of insulin can cause a significant decrease in blood glucose concentrations in anaesthetized and mechanically ventilated pigs and results in clinically relevant serum insulin levels. Similar effects in humans would make inhaled insulin possible for clinical use.