Purpose: To study the short-term reproducibility of lung density measurements by multislice computed tomography (CT) using three different radiation doses and three reconstruction algorithms. Material and Methods: Twenty-five patients with smoker's emphysema and 25 patients with α₁rantitrypsin deficiency underwent 3 scans at 2-week intervals. Low-dose protocol was applied, and images were reconstructed with bone, detail, and soft algorithms. Total lung volume (TLV), 15th percentile density (PD-15), and relative area at -910 Hounsfield units (RA-910) were obtained from the images using Pulmo-CMS software. Reproducibility of PD-15 and RA-910 and the influence of radiation dose, reconstruction algorithm, and type of emphysema were then analysed. Results: The overall coefficient of variation of volume adjusted PD-15 for all combinations of radiation dose and reconstruction algorithm was 3.7%. The overall standard deviation of volume-adjusted RA-910 was 1.7% (corresponding to a coefficient of variation of 6.8%). Radiation dose, reconstruction algorithm, and type of emphysema had no significant influence on the reproducibility of PD-15 and RA-910. However, bone algorithm and very low radiation dose result in overestimation of the extent of emphysema. Conclusion: Lung density measurement by CT is a sensitive marker for quantitating both subtypes of emphysema. A CT-protocol with radiation dose down to 16 mAs and soft or detail reconstruction algorithm is recommended.