A common method for analyzing a multibreath nitrogen washout (MBNW) is to perform moment analysis and derive the mean dilution number (MDN). A homogeneously mixed alveolar space with zero series dead space (VD = 0) will always result in a MDN = 1, regardless of breathing pattern. A higher MDN implies more inhomogeneity. But, if VD greater than 0, the MDN can become sensitive (artificially high) to VD/VT ratios. We present an alveolar-based mean dilution number (AMDN) that uses the cumulative expired alveolar volume. Unlike the MDN, the AMDN for a homogeneously mixed alveolar space is unity, regardless of VD or VT, and hence should be a more appropriate index of inhomogeneity at the alveolar level. Two sets of experiments were used to compare the AMDN with the MDN. First, a MBNW was performed by five healthy subjects at spontaneous VD/VT and at a low VD/VT achieved by a controlled increase in VT. Here, the MDN decreased from 1.98 +/- 0.1 to 1.79 +/- 0.06, whereas the AMDN was essentially unchanged (1.42 +/- 0.04 to 1.38 +/- 0.06). Second, MBNW values from seven healthy subjects, five with cystic fibrosis, and 10 asthmatic subjects (before and after bronchodilation) were analyzed. Compared with the MDN, the AMDN showed a significantly wider separation between clinical groups. Also, the AMDN demonstrated an increased variability within both sick groups versus a decrease in the healthy group. We conclude that the AMDN is superior to the MDN because of its decreased sensitivity to breathing pattern but increased sensitivity to degree of disease.