BACKGROUND AND METHOD: Despite significant but similar degree of expiratory air-flow limitation, wide ranges of PaCO2 can be seen in chronic obstructive pulmonary disease (COPD), and it is questionable whether analysis of only expiratory parameters on pulmonary function test (PFT) is enough to estimate the severity of COPD and can predict the CO2 retention. To evaluate the usefulness of inspiratory parameters of PFT to predict CO2 retention, the author conducted pulmonary function test (FVC maneuver) and arterial blood gas analysis concurrently in 32 patients with COPD suffering from severe degree of expiratory airflow limitation (FEV1< or =50% of predicted value). RESULT: Between CO2 retainers (PaCO2>42 mmHg, N=16) and Non-retainers (PaCO2< or =42mmHg, N=16), there were no significant differences in their age, sex distribution, FVC, FEV1, FIVC, and PaO2. Peak inspiratory flow(PIF) was significantly lower in CO2 retainers than in non-retainers (49.5+/-2.4 vs. 88.8+/-25.4% of predicted value, p<0.001), and FEV1/FVC (45.1+/-6.1 vs. 54.7+/-10.7, p<0.01) also. PIF (r=-0.5431, p<0.01), FEV1/FVC (R=-.3346, p<0.05), FEV1 (r=-.2983, p<0.05) showed significant inverse correlation with PaCO2. Among the above parameters, PIF was the only independent parameters that could determine PaCO2 by the stepwise multiple regression analysis. A prediction equation of PaCO2 on this base shows that PaCO2 = -0.16756PIF + 55.061162 CONCLUSION: It was suggested that PIF reflecting the airflow limitation during the inspiratory phase has an important contribution to the CO2 retention as well as some expiratory parameters such as FEV1, FEV1/FVC.