Theophylline disposition (clearance, volume of distribution, and half-life) was measured in 31 normal volunteers and 26 acutely ill patients with airway obstruction within 24 hours of hospital admission and again when intravenous aminophylline was discontinued 2 to 10 days later. Sex, age, race, the diagnosis of asthma, and the diagnosis of chronic bronchitis were not significantly correlated with clearance. Clearance was significantly increased in smokers and significantly decreased in patients with congestive heart failure, pneumonia, and severe bronchial obstruction. The volume of distribution was not correlated with these variables. Within 24 hours of admission, the mean clearance in our patients without congestive heart failure or pneumonia was 44.5 ml per hour per kg of body weight, 40 per cent less than the value on which the widely used 0.9 mg per kg per hour dosage is based. The unexplained interindividual variability of clearance was significantly higher amont patients than control subjects, but intraindividual variability in clearance was not different in these 2 groups after correction for the effects of smoking, congestive heart failure, and so on. Large variability in clearance among patients makes the plasma theophylline concentration resulting from any dosage relatively uncertain. Based on the the results of this study, recommendations are made concerning theophylline dosage computation, and the clinical circumstances calling for determination of plasma theophylline concentrations.