The pulmonary deposition and pharmacokinetics of insulin, administered via an endotracheal tube as an aerosol and instillate, in formulations containing either 113mIn-DTPA or 99mTc-DTPA (for gamma scintigraphic imaging) have been studied in four male New Zealand White rabbits. Using a randomized crossover design, the pharmacokinetics of intravenous insulin were also characterized. Recovery of immunoreactive insulin after nebulization was greater than 90%, indicating that the aerosolisation procedure did not cause appreciable insulin degradation. Gamma scintigraphy demonstrated that the penetration index (peripheral:central deposition) for the aerosolized formulation (1.52) was much greater than that for the instillate (0.32). Gamma scintigraphy also allowed exact quantification of the dose deposited after aerosol administration and thus permitted accurate determination of bioavailabilities. The bioavailable fraction for aerosolized insulin was 10-fold greater than for instilled insulin (57.2 vs 5.6%). Mucociliary clearance was likely to be greater for the instillate since it showed a preferential central deposition; this may account for the lower bioavailability. Insulin pharmacokinetics from both pulmonary formulations were absorption rate limited, resulting in postpeak half-lives which were approximately 20-fold greater than the intravenous elimination half-life (3 min). The apparent absorption rate constants resulting from instillation and aerosolisation were statistically equivalent (0.015 and 0.011 min-1, respectively). Mucociliary clearance of insulin would result in an overestimation of the true absorption rate constant; hence if mucociliary transport were greater for the instillate, then the true airways to blood transfer rate constant will be higher for the aerosolized formulation.