The measurement and interpretation of "diffusing capacity" by either single-breath or steady-state methods are complicated by both technical and conceptual difficulties. The CO uptake fraction is less complex but, as originally described, it is unacceptably sensitive to dead-space ventilation. A modification (the "alveolar CO uptake fraction", UA) largely removes this factor. We have measured UA in thirteen healthy subjects and 100 patients with a variety of pulmonary disorders. It is reproducible and appears sensitive to clinical abnormality. Its technical and interpretative simplicity suggest its use as an alternative to other measures of CO transfer.