Assessment in young children of the bioavailability of slow-release theophylline formulations is hampered by the requirement for frequent blood sampling. Calculations of bioavailability from serial serum and passively absorbed saliva samples were therefore compared in six 9- to 12-year-old asthmatic children receiving multiple doses of Theo-Dur Sprinkle every 12 hours, using Theo-Dur tablets, a previously characterized formulation, as a reference. Results indicated 85 +/- 5 percent and 82 +/- 8 percent (mean +/- SEM) relative bioavailability based on serum and salivary measurements, respectively. Correlation coefficient for serum and passively absorbed saliva bioavailabilities was 0.90. Passively absorbed saliva provides an acceptably accurate, noninvasive method for theophylline bioavailability assessment and may be a useful alternative for bioavailability studies in young children.