In patients with insulin-dependent diabetes mellitus (IDDM) isolated peripheral airway involvement may give rise to inspiratory threshold load (ITL) contributing to dyspnea. Based on the reported evidence of a greater increase in end-expiratory lung volume (EELV) with hypoxia than with hypercapnia in IDDM, we wondered whether, and to what extent in the two conditions, EELV contribute to perception of dyspnea (PD). We studied five nonsmokers aged between 19 and 45, with IDDM under good metabolic control and five normal control subjects matched for age. In each patient, we evaluated the electromyographic activity of the diaphragm (Edi), the swings of esophageal (Pessw), gastric (Pgsw), and transdiaphragmatic (Pdisw = Pgsw-Pessw) pressures; PD was assessed by a modified Borg scale during hypercapnic-hyperoxic (HCH) and hypoxic-isocapnic (HIC) stimulation. Change in inspiratory capacity (IC) was considered the mirror image of increase in EELV, that is, dynamic hyperinflation (DH), while intrinsic positive end inspiratory pressure (PEEPi) was measured as an index of inspiratory threshold load (ITL). In controls, Edi and Pdi but not their ratio (Edi/Pdi) related to Borg. In patients the following was found: (1) with each of the two stimuli, for any given Edi, Pdi, and Edi/Pdi ratio, there was greater Borg than in controls, (2) a similar increase in ITL and DH with HCH and HIC, (3) Edi/Pdi related to Borg similarly with HCH as with HIC. In conclusion, in controls, Edi and Pdi were associated with the perception of dyspnea similarly with the two chemical stimuli. In this subset of patients with IDDM, Edi/Pdi ratio throughout increase in EELV and ITL was found to affect the perception of dyspnea in hypoxia to a similar extent as in hypercapnia.