The adverse effect of long-term steroid treatment on respiratory muscle function remains controversial. We evaluated inspiratory muscle strength and endurance in steroid-dependent asthmatics in comparison with other asthmatics or with patients with chronic obstructive pulmonary disease exhibiting a comparable level of lung hyperinflation. Inspiratory muscle function was assessed by maximal inspiratory pressure (Pimax) and by an incremental inspiratory threshold loading test in 19 patients who had had steroid-dependent asthma (SDA) requiring a mean daily dose of 20.7 +/- 0.8 mg prednisone for 5 +/- 1.4 yr. They were compared with 16 healthy control subjects, 30 patients with COPD, and 16 patients with non-steroid-dependent asthma (NSDA). Pimax as percentage of predicted values (%Pimax) was not significantly different in patients with SDA (77 +/- 5%) or NSDA (83 +/- 6%) than in control subjects (93 +/- 4%). In contrast, %Pimax was lower in patients with COPD (59 +/- 4.4%) than in those with SDA or NSDA (p < 0.05) or the control subjects (p < 0.0001). A significant correlation was found between %Pimax and hyperinflation assessed by the FRC/TLC ratio (r = 0.42; p < 0.001). Inspiratory endurance, defined as the ratio of maximal peak inspiratory pressure sustained for 2 min to individual Pimax (Plim2/Pimax), was significantly lower in the SDA (43 +/- 3%; p < 0.0001), NSDA (65 +/- 4%; p = 0.01), and COPD (55 +/- 3%; p < 0.0001) groups than in the control group (76 +/- 2%). Plim2/Pimax was also lower in patients with SDA than in those with COPD (p = 0.0073) or NSDA (p < 0.0001). Hyperinflation plays a major role in inspiratory muscle dysfunction associated with obstructive lung disorders, but the finding of a significantly decreased endurance in patients with SDA when compared with patients with COPD, despite a lower level of hyperinflation in the former group, points to a deleterious effect of long-term corticosteroid treatment on inspiratory muscle function in asthmatics.