There is considerable intersubject variability in the perceived intensity of breathlessness for a given level of activity among patients with chronic airflow limitation (CAL). To examine possible factors contributing to this variability we compared breathing pattern parameters, dynamic operational lung volumes, and Borg dyspnea ratings in 23 patients with severe CAL and in 10 healthy age-matched normal subjects during cycle ergometry to symptom-limitation. Patients with CAL had significantly (p < 0.01) higher levels of ventilation (% maximal voluntary ventilation) for a given work rate (slope of V̇e(%MVV)/WR(% pred max) = 1.51 ± 0.18 versus 0.63 ± 0.10; mean ± SEM) and greater dynamic lung hyperinflation (DH) (change [Δ] in end-expiratory lung volume [EELVdyn] = +0.31 ± 0.11 L versus −0.16 ± 0.22 L). Compared with normal subjects at a standardized V̇e (30 L/min), the CAL group was more breathless (Borg = 4 ± 1 versus 2 ± 1, p < 0.01) and hyperinflated (EELVdyn = 75 ± 3 versus 46 ± 6 %TLC, p < 0.001; end-inspiratory lung volume [EILVdyn] = 85 ± 3 versus 67 ± 5 %TLC, p < 0.01). Within the CAL group, change in Borg ratings correlated with ΔV̇e(%MVV) (r = 0.77, p < 0.001) and with slope of V̇e(%MVV)/WR(% pred max) (r = 0.48, p < 0.01). Regression analysis selected ΔEILVdyn (or Δ inspiratory reserve volume [ΔIRVdyn]) from various dynamic ventilatory parameters as the strongest predictor of Δ Borg (r = 0.63, p < 0.001). Components of ΔEILVdyn (i.e., ΔEELVdyn, ΔVt) each contributed significantly (p < 0.001) to breathlessness and with A breathing frequency accounted for 61% of the variance in ΔBorg (r = 0.78, f = 38.20, p < 0.001). Accounting for ventilation, EELVdyn continued to contribute significantly to breathlessness; at a standardized V̇e (30 L/min), EELVdyn predicted 31% of the variance in Borg ratings (p < 0.01). Exertional breathlessness in CAL is a function of ventilatory demand and intensifies with encroachment on the inspiratory reserve volume or ventilatory reserve. Acute DH, with its attendant intrinsic mechanical loading, contributes importantly to intersubject variability in the perception of breathlessness for a given ventilation.