In a double-blind controlled trial 14 chronic asthmatic patients with regular nocturnal exacerbations took 16 mg slow-release oral salbutamol (two Ventolin spandets), 450 mg slow-release aminophylline (two Phyllocontin Continus tablets), or placebo at midnight. Mean peak expiratory flow rates on waking were significantly higher on the active drugs than on placebo (p < 0.01 for salbutamol; p < 0.05 for aminophylline) but neither drug abolished the overnight fall in PEFR. Plasma drug levels at 0600 hr were 17.3 ng/ml (+/- 5.3 ng/ml SD) for salbutamol, and 7.1 micrograms/ml (+/- 3.1 micrograms/ml SD) for theophylline. Steady-state derived from plasma levels of salbutamol during intravenous infusion indicated that the morning salbutamol levels were probably in a therapeutic range for asthma. The morning theophylline levels, however, were suboptimal when aminophylline was given only at night.