The study investigated whether high-intensity exercise impairs inspiratory and expiratory muscle perfusion in patients with COPD. We compared respiratory local muscle perfusion between constant-load cycling (sustained at 80% WRpeak) and voluntary normocapnic hyperpnoea reproducing similar work of breathing (WoB) in 18 patients (FEV1:58±24% predicted). Local muscle blood flow index (BFI), using indocyanine green dye and fractional oxygen saturation (%StiO2)were simultaneously assessed by near-infrared spectroscopy (NIRS) over the intercostal, scalene, rectus abdominis and vastus lateralis muscles. Cardiac output (impedance cardiography), WoB (oesophageal/gastric balloon catheter), and diaphragmatic and extradiaphragmatic respiratory muscle electromyographic activity (EMG) were also assessed throughout cycling and hyperpnoea. Minute ventilation, breathing pattern, WoB and respiratory muscle EMG were comparable between cycling and hyperpnoea. During cycling, cardiac output and vastus lateralis BFI were significantly greater compared to hyperpnoea [by +4.2(2.6-5.9) L/min and +4.9(2.2-7.8) nmol/s], respectively, (p<0.01). Muscle BFI and %StiO2 were respectively lower during cycling compared to hyperpnoea in scalene [by -3.8(-6.4- -1.2) nmol/s and -6.6(-8.2- -5.1)%], intercostal [by -1.4(-2.4- -0.4) nmol/s and -6.0(-8.6- -3.3)%] and abdominal muscles [by -1.9(-2.9- -0.8) nmol/s and -6.3(-9.1- -3.4)%] (p<0.001). The difference in respiratory (scalene and intercostal) muscle BFI between cycling and hyperpnoea was associated with greater dyspnoea (Borg CR10) scores (r= -0.54 and r= -0.49, respectively, p<0.05). These results suggest that in patients with COPD 1) locomotor muscle work during high-intensity exercise impairs extradiaphragmatic respiratory muscle perfusion and that 2) insufficient adjustment in extradiaphragmatic respiratory muscle perfusion during high-intensity exercise may partly explain the increased sensations of dyspnoea.