The pharmacokinetics of intranasal insulin containing a medium-chain phospholipid (didecanoyl-L-alpha-phosphatidylcholine) as absorption enhancer, was studied in normal volunteers by measuring plasma glucose, insulin, C-peptide, and glucagon. Eleven fasting subjects received 4 U insulin intravenously, 6 U subcutaneously, or three doses intranasally (approximately 0.3 U kg-1, 0.6 U kg-1, 0.8 U kg-1) in random order on five separate days. Intranasal insulin was absorbed in a dose-dependent manner with a mean plasma insulin peak 23 +/- 7 (+/- SE) min after administration. Mean plasma glucose nadir was seen after 44 +/- 6 min, 20 min later than following intravenous injection. Furthermore, intranasal administration of insulin resulted in a faster time-course of absorption than subcutaneous injection, with significantly reduced intersubject variation (p less than 0.001). Bioavailability for the nasal formulation was 8.3% relative to an intravenous bolus injection when plasma insulin was corrected for endogenous insulin production estimated by C-peptide. A dose-dependent suppression of C-peptide and stimulation of glucagon secretion occurred after intranasal administration of insulin. Nasal irritation from spraying was absent or slight.