Purpose: We generated lung morphometry measurements using single-breath diffusion-weighted MRI and three different acinar duct models in healthy participants and patients with emphysema stemming from chronic obstructive lung disease (COPD) and alpha-1 antitrypsin deficiency (AATD). Methods: Single-breath-inhaled 3 He MRI with five diffusion sensitizations (b-value ¼ 0, 1.6, 3.2, 4.8, and 6.4 s/cm 2 ) was used, and signal intensities were fit using a cylindrical and single-compartment acinar-duct model to estimate MRIderived mean linear intercept (L m ) and surface-to-volume ratio (S/V). A stretched exponential model was also developed to estimate the mean airway length and L m . Results: We evaluated 42 participants, including 15 elderly never-smokers (69 6 5 years), 12 ex-smokers without COPD (67 6 11 years), 9 COPD ex-smokers (80 6 6 years), and 6 AATD patients (59 6 6 years). In the never-and ex-smokers, the diffusing capacity of the lung for carbon monoxide (DL CO ) and computed tomography relative area of less than 2950 Hounsfield units (RA 950 ) were normal, but these were abnormal in the COPD and AATD patients, which is reflective of emphysema. Although cylindrical and stretched-exponential-model estimates of L m and S/V were not significantly different, the single-compartment-model estimates were significantly different (P < 0.05) for the never-and ex-smoker subgroups. All models estimated significantly worse L m and S/V in the AATD and COPD subgroups compared with the never-and exsmokers without emphysema. Conclusions: Differences in airspace enlargement may be estimated using L m and S/V, generated using MRI and a stretchedexponential or cylindrical model of the acinar ducts. Magn Reson Med 79:439-448,