The effects of diurnal variation on bioavailability assessments were examined using computer-simulated data based on the changes observed in theophylline kinetics. During one 12-hour dosage interval (noon to midnight), clearance was assumed to be larger than during the other dosage interval (midnight to noon). Oral data was simulated until steadystate occurred. Intravenous bolus data, which represented a stable-isotope pulse dose, was also simulated for both the high and low clearance dosage intervals. When the respective areas under the serum concentration-time curves were compared, the systemic availability (F = AUCpo/AUCiv) during the dosage interval with the larger clearance was greater than 1.0, but during the dosage interval with the smaller clearance it was less than 1.0. When computing the bioavailability of a drug, diurnal variations should be assessed as a potential cause of variation.