OBJECTIVE To evaluate the intratracheal route of administration as an alternative to oral administration for 2',3'-dideoxyinosine (ddI). METHODS A ddI dose (40 mg/kg/300 microliters or 6.5 mg/kg/50 microliters) was instilled into the trachea in female Fisher rats and an intravenous tracer dose (9 micrograms/kg) of 3H-ddI was administered concomitantly to determine the drug clearance. Plasma concentrations were analyzed for the rate and extent of absorption. RESULTS ddI was rapidly absorbed from the lungs, with a bioavailability of 63% at 40 mg/kg and 101% at 6.5 mg/kg. By comparison, our previous data showed an oral bioavailability of about 15% (Pharm Res., 9:822, 1992). The distribution of a dye solution instilled intratracheally showed that a fraction of the 300 microliters dose spilled over to the gastrointestinal tract, where the entire 50 microliters dose was retained in the lungs. The different distribution of the two doses volumes likely contributed to the different bioavailability, with a fraction of the higher dose/volume degraded in the gastrointestinal tract after the spillover. Absorption of ddI from the airspace of the lung was biexponential, suggesting two absorption processes. CONCLUSIONS These data indicate significantly higher and less variable bioavailability of ddI by the intratracheal route of delivery compared to the oral route. Furthermore, the complete bioavailability at the lower dose/volume indicates no significant pulmonary first pass elimination for ddI.