In clinical practice, an elevated carbon monoxide (CO) transfer coefficient (KCO) and restrictive ventilatory defect are taken as features of respiratory muscle weakness (RMW). However, the authors hypothesised that both pattern and severity of RMW effect gas transfer and lung volumes.Measurements of CO transfer and lung volumes were performed in patients with isolated diaphragm weakness (n=10), inspiratory muscle weakness (n=12), combined inspiratory and expiratory muscle weakness (n=5) and healthy controls (n=6).Patients with diaphragm weakness and inspiratory muscle weakness had reduced total lung capacity (TLC) (83.6% predicted and 68.9% pred, respectively), functional residual capacity (FRC) (83.9% pred and 83.6% pred) and transfer factor of the lung for CO (TL,CO) (86.2% pred and 66.2% pred) with increased KCO (114.1% pred and 130.2% pred). Patients with combined inspiratory and expiratory muscle weakness had reduced TLC (80.9% pred) but increased FRC (109.9% pred) and RV (157.4% pred) with decreased TL,CO (58.0% pred) and KCO (85.5% pred).In patients with diaphragm weakness, the increase in carbon monoxide transfer coefficient was similar to that of normal subjects when alveolar volume was reduced. However, the increase in carbon monoxide transfer coefficient in inspiratory muscle weakness was often less than expected, while in combined inspiratory and expiratory muscle weakness, the carbon monoxide transfer coefficient was normal/reduced despite further reductions in alveolar volume, which may indicate subtle abnormalities of the lung parenchyma or pulmonary vasculature. Thus, this study demonstrates the limitations of using carbon monoxide transfer coefficient in the diagnosis of respiratory muscle weakness, particularly if no account is taken of the alveolar volume at which the carbon monoxide transfer coefficient is made. Transfer factor of the lung for carbon monoxide (CO) (TL,CO) is the product of the CO transfer coefficient (KCO), which indicates the rate constant of CO uptake from the alveoli and the alveolar volume (VA) at which the measurement is made [1]. Many studies, including the original study by KROGH [1], have shown that if CO transfer is measured at submaximal VA, the decline in TL,CO is partially offset by an increase in KCO. In clinical practice, CO uptake is measured at total lung capacity (TLC), but many disease states associated with a reduced TLC often have coexistent intrapulmonary disease and, as a consequence, have a reduction in KCO. However, when the cause of the reduced TLC is extrapulmonary, a distinctive pattern may be found in which KCO is increased at full inflation and VA is decreased, so that reductions in TL,CO are smaller than observed in intrapulmonary disease. The most obvious disorder simulating the effects of submaximal inflation is inspiratory muscle weakness without intrapulmonary disease. In an influential study of six patients [2], in which TLC was reduced due to severe isolated diaphragm weakness, a relatively benign disease in the absence of other...