Objective: To observe the relationship of deep slow respiratory pattern and respiratory impedance(RI) in patients with chronic obstructive pulmonary disease (COPD). Methods: RI under normal respiration and during deep slow respira tion was measured one after the other with impulse oscillometry for 8 patients with COPD and for 9 healthy volunteers as control. Results: When r espiration was changed from normal pattern to the deep slow pattern, the tida l volume increased and respiratory frequency significantly decreased in both gro ups , the total respiratory impedance (Z respir) showed a decreasing trend in COPD group, but with no obvious change in the control group. No chang e in the resonant frequency (fres) was found in both groups, and the respiratory viscous resistance obviously decreased in the COPD group(R5: P =0.0168 ; R20: P =0.0498; R5—R20: P =0.0388),though in the control group it was unchanged. Conclusion: IOS detection could reflect the response he terogeneity of different compartments of respiratory system during tidal breathi ng. During deep slow respiration, the viscous resistance in both central airw ay and peripheral airway was decreased in patients with COPD. RI measurement by impulse oscillometry may be a convenient pathophysiological method for studying the application of breathing exercise in patients with COPD. Objective: To observe the relationship of deep slow respiratory pattern and respiratory impedance (RI) in patients with chronic obstructive pulmonary disease (COPD). Methods: RI under normal respiration and during deep slow respiration was measured one after the other with impulse oscillometry for 8 patients with COPD and for 9 healthy volunteers as control. Results: When r espiration was changed from normal pattern to the deep slow pattern, the tida l volume increased and respiratory frequency significantly decreased both gro ups, the total respiratory impedance (Z respir ) showed a decreasing trend in COPD group, but with no obvious change in the control group. No chang e in the resonant frequency (fres) was found in both groups, and the respiratory viscous resistance obviously decreased in the COPD group (R5: P = 0.0168; R20: P = 0.0498; R5-R20: P = 0.0388), though in the control group it was unchanged. Conclusion: IOS detection could reflect the response he tergeneity of different compartments of respiratory system during tidal breathi ng. During deep slow respiration, the viscous resistance in both central airw ay and peripheral airway was decreased in patients with COPD. RI measurement by impulse oscillometry may be a convenient pathophysiological method for studying the application of breathing exercise in patients with COPD.