Ten normal subjects ingested lithium carbonate (600 mg p.o. b.i.d.) for three 1-week intervals. At the end of each weekly interval, subjects' lithium clearances were randomly perturbed for 12 hours with the subject in a supine position by infusing either normal saline (308 mEq), sodium bicarbonate (350 mEq) in normal saline (308 mEq), or theophylline (mean = 14.0 micrograms/ml) in normal saline (308 mEq). Subjects were placed on a 200 mEq/day sodium diet during the lithium clearance perturbation stages of the study. When each patient's normal saline lithium clearance was used as a control, it was found that the theophylline produced a significantly greater % increase in lithium clearance than did the sodium bicarbonate. Theophylline infusions increased patients' individual lithium clearances by 51 +/- 52%, while sodium bicarbonate infusions increased lithium clearances by 0.6 +/- 33%. Theophylline infusions ought to be investigated as an alternative to hemodialysis in lithium intoxications requiring the immediate reduction of lithium concentrations.