In normal seated subjects we increased single-breath ventilation inhomogeneity by changing both the preinspiratory lung volume and breath-hold time and examined the ensuing effects on two different techniques of measuring the diffusing capacity of the lung for carbon monoxide (DLCO). We measured the mean single-breath DLCO using the three-equation method (DLCOSB-3EQ) and also measured DLCO over discrete intervals during exhalation by the "intrabreath" method (DLCOexhaled). We assessed the distribution of ventilation using the normalized phase III slope for helium (SN). DLCOSB-3EQ was unaffected by preinspiratory lung volume and breath-hold time. DLCOexhaled increased with increasing preinspiratory lung volume and decreased with increasing breath-hold time. These changes correlated with the simultaneously observed changes in ventilation inhomogeneity as measured by SN (P < 0.01). We conclude that measurements of DLCOexhaled do not accurately reflect the mean DLCO. Intrabreath methods of measuring DLCO are based on the slope of the exhaled CO concentration curve, which is affected by both ventilation and diffusion inhomogeneities. Although DLCOexhaled may theoretically provide information about the distribution of CO uptake, the concomitant effects of ventilation nonuniformity on DLCOexhaled may mimic or mask the effects of diffusion nonuniformity.