When testing the response to beta 2-agonist drugs in severe chronic obstructive pulmonary disease (COPD), a dose-response assessment should be undertaken. This study compares the time course of inhaled salmeterol (25, 50 and 75 micrograms) and formoterol (12, 24 and 36 micrograms) at different doses in a group of 12 patients with partially reversible, but severe COPD (FEV1 of 12-32% of predicted values after beta 2-agonist drugs had been withheld for 24 h). All doses of salmeterol and formoterol induced a significant (P < 0.01) spirometric improvement over the 12-h monitoring period, when compared to the spirometric improvement after placebo, but while formoterol induced a dose-dependent increase of the FVC, FEV1 and FEF50, this was not the case for salmeterol. In fact, 75 micrograms salmeterol did not produce a further improvement of these parameters. Mean peak bronchodilation, expressed as the increase in FEV1 over baseline values, occurred 2 h after inhalation of the three doses of salmeterol, and 1 h after inhalation of the three doses of formoterol. A comparison of 50 micrograms salmeterol with 12 micrograms or 24 micrograms formoterol (clinically recommended doses), showed that improvement of FEV1 after salmeterol was statistically (P < 0.05) higher than that after the two doses of formoterol, although the mean peak bronchodilations were similar. This was because salmeterol has a longer duration of action than formoterol. These data demonstrate that salmeterol is equally effective as, but longer-acting than, formoterol at clinically recommended doses in patients suffering from COPD, with severe airway obstruction.(ABSTRACT TRUNCATED AT 250 WORDS)