The population pharmacokinetic parameters for a once-daily administered preparation, Uniphyl were estimated from data collected in the premarketing clinical trial. Altogether, 2772 serum theophylline concentrations were obtained from 131 normal subjects and 306 patients suffering from chronic asthma or chronic obstructive pulmonary disease who participated in the phase I, II, and III clinical trials in Japan. The serum concentration profile was described by a linear one-compartment model with first-order absorption. The factors affecting the pharmacokinetics of this drug were examined by the likelihood ratio test using a nonlinear mixed effect model (NONMEM). The first-order absorption rate constant (Ka) for a 200-mg tablet in a fasting condition was obtained as 0.0773 (1/h), which was smaller than the elimination rate constant (0.168 1/h), indicating the flip-flop characteristic of this preparation. Food ingestion increased the Ka by 17% and the absorption lag time by 5-fold but did not affect the extent of absorption. The 400-mg tablet showed a Ka value 19%, smaller than the 200-mg tablet. Children not older than 15 years showed 58% longer absorption lag time. The inter-individual variability in Ka was 19%, suggesting small variability in the in vivo release process. The total body clearance was related to hepatic function, smoking habits, and age. Furthermore, clearance decreased in association with the severity of illness. The findings obtained here are useful not only for the initial dosage adjustment for patients with a variety of backgrounds but also for dose individualization based on serum concentration monitoring with or without the Bayesian feedback method.