This study was designed to test whether single oral doses of theophylline, rather than the specified single intravenous infusions, could be used with a nomogram designed to predict the maintenance oral dose necessary to establish plasma theophylline concentrations of 10 micrograms ml (Koup et al., 1979) in patients with obstructive airways disease. A test dose of theophylline (5 mg kg) was administered on two separate occasions as an oral elixir and as an intravenous infusion. Exactly 6 h after each test dose, plasma theophylline concentration was measured, and the result used to predict daily maintenance dose requirements. During subsequent oral administration, a sustained-release preparation was given in equally divided 12-hourly doses, and 'steady-state' concentrations were measured at trough. Nineteen patients completed the study, and on doses ranging from 4.2 to 21.1 mg kg-1 day-1, trough concentrations of 5.4 to 15.8 micrograms ml (mean 10.0 +/- 0.7 micrograms ml) were achieved. No significant differences were noted between the dose predictions made following either the oral or the intravenous test doses, and the predictive accuracy using the oral route was not compromised. This would suggest that the test dose of theophylline may be administered orally, instead of intravenously, with this nomogram without reducing its efficiency.