This study investigated the hypothesis that hypercapnia in some chronic obstructive pulmonary disease (COPD) patients may be related to a high cerebrovascular response to carbon dioxide (CO2). The relationship between responses of ventilation and of cerebral blood volume (CBV) to acute changes in carbon dioxide tension in arterial blood (Pa,CO2) was measured in 17 chronic hypercapnic (Pa,CO2 >6.0 kPa) and 16 normocapnic (Pa,CO2 < or = 6.0 kPa) COPD patients, who were matched for degree of airway obstruction (forced expiratory volume in one second 27% predicted). Results were compared with 15 age-matched healthy subjects. CBV was measured using near infrared spectroscopy during normo- and hypercapnia and related to inspired minute ventilation (V'I) and mouth occlusion pressure (P0.1). Hypercapnia (end-tidal pressure of carbon dioxide (deltaPET,CO2) > 1 kPa) was induced by giving adequate amounts of CO2 in the inspired air. During normocapnia, CBV (mL x 100 g(-1)) was 2.41+/- 0.66 and 2.90 +/- 0.60 (mean +/- SD) in the normocapnic and chronic hypercapnic patients, respectively, which was significantly lower compared to healthy subjects (3.53 +/- 0.77). All slopes of CO2 responsiveness (deltaCBV/deltaPa,CO2, deltaV'I/deltaPa,CO2, deltaP0.1/deltaPa,CO2) were significantly lower in both COPD groups relative to healthy subjects, but were not significantly different between the COPD groups. A poor but positive correlation between ventilatory and cerebrovascular CO2 responsiveness (deltaCBV/deltaPa,CO2 and deltaV'I/deltaPa,CO2) was found in COPD patients and healthy subjects. The findings do not support the hypothesis of abnormal cerebrovascular responses to carbon dioxide in hypercapnic chronic obstructive pulmonary disease patients.