Carbon dioxide (CO2) retention occurs in some but not all patients with obstructive pulmonary disease. In order to assess if the pattern of ventilation modulates CO2 retention, 15 normocapnic (group 1) and 15 hypercapnic (group 2) patients with severe chronic obstructive pulmonary disease (forced expiratory volume in 1 second (FEV1) less than or equal to 1.0 liter) were studied retrospectively. Utilizing clinical information, anion gap and acid-base nomogram, subjects with superimposed acid-base disturbances were eliminated. Therefore, only patients who exhibited steady state ventilatory patterns were studied. In group 1, mean arterial carbon dioxide tension (PaCO2) was 40 ł 2 torr and mean arterial pH (pHa) was 7.410 ł0.004. In group 2, mean PaCO2 was 52.5 ł1.2 torr and pHa was 7.390 ł0.007. No statistically significant differences between groups were present with respect to age. height, sex, lung volumes and flow rates, diffusing capacity and CO2 production. Minute ventilation was similar in both groups (7.631 liters and 7.81 liters). In group 2, the patients had a significantly higher respiratory rate per minute (22 versus 16.5) and smaller tidal volume (355 versus 463 cc) than the patients in group 1. This pattern of ventilation resulted in a larger dead space ventilation (3.98 liters versus 2.95 liters) and a lower alveolar ventilation (3.82 liters versus 4.61 liters) with consequent CO2 retention. The higher respiratory frequency in the patients in group 2 may be due to vagal stimulation from the lungs since this group had a fivefold greater incidence of chronic bronchitis and a seven-fold greater incidence of cor pulmonale than the patients in group 1.