Objective: To evaluate the value of single-source dual-energy CT (ssDECT) in differentiating lipid-poor adenomas from metastases in adrenal glands. Methods: From August 2011 to Oct 2014, 63 patients with 73 adrenal nodules (53 metastases proven by 5-6 months follow-up, and 20 histopathologically proven adenomas, CT value > 10 HU) underwent ssDECT scanning. The CT values of conventional polychromatic CT and virtual monochromatic images (40-140 keV) , fat-water density images and effective atomic number (eff-Z) were reconstructed on an ADW4.5 workstation and ROC curves were then constructed to evaluate the effectiveness of each parameter. The slope of spectral curve was measured and divided into 3 types: increment curve (K>0.1), straight curve (-0.1≤K≤0.1) and decrement curve (K< -0.1) according to the slope (the value of K) of spectral curve, and the curve patterns in the two groups were compared statistically. Results: There was no statistical difference between the mean CT values of metastases (35.12±5.29)HU and lipid-poor adenomas (32.48±6.94)HU by conventional polychromatic CT (P>0.05). The range of single-energy CT values of metastases [from (53.00±15.12) HU to (33.38±5.67) HU] was significantly higher than that of lipid-poor adenomas [from (26.90±26.94) HU to (28.77±10.66) HU] at energy levels ranging from 40 to 80 keV (P<0.05). There was no significant difference between the single-energy CT value of metastases and lipid-poor adenomas at energy levels ranging from 90 to 140 keV (P>0.05). The median fat-water concentration of metastases was -164.61 μg/cm3, significantly lower than that of lipid-poor adenomas (114.32 μg/cm3,P<0.05). The eff-Z of metastases (7.76±0.15) was also significantly higher than that of lipid-poor adenomas (7.50±0.25, P<0.05). When the threshold of fat-water concentration was -143.89 μg/cm3, the sensitivity, specificity and accuracy rate of metastasis diagnosis was 70.0%, 66.0%, and 76.7%, respectively. When the threshold of eff-Z was 7.63, the sensitivity, specificity and accuracy rate of metastasis diagnosis was 83.0%, 65.0%, and 80.4%, respectively. The lower the energy, the higher the diagnostic accuracy at energy levels ranging from 40 to 80 keV, and that of 40 keV was the highest. The spectral curves of metastases included 2 (3.8%) ascending curves, 9 (17.0%) straight curves and 42 (79.2%) descending curves, while in the 20 lipid-poor adenomas, there were 9 (45.0%) ascending curves, 4 (20.0%) straight curves and 7 (35.0%) descending curves, showing significant differences between the two groups (P<0.05). Conclusions: Single-source dual-energy CT provides an effective multi-parameter approach for differentiating lipid-poor adrenal adenomas from metastases.