Objective To explore the role of high-resolution computed tomography (HRCT) phenotype and low-attenuation area (LAA) score system in the comprehensive assessment of severity of COPD. Methods A total of 197 COPD patients (stable stage), including 162 males and 35 females, who were admitted in Zhujiang Hospital, Southern Medical University during June 2014 to December 2016 were involved in this study. Basic information, pulmonary function, respiratory muscle strength, symptoms and quality of life, exercise capacity, nutritional status,anxiety and depression, and HRCT scans were recorded. Patients were classified as phenotype A, E and M based on HRC scan. Correlation analysis of clinical regression with LAA score or phenotypes were performed. Results 197 subjects were divided into three phenotypes based on HRCT scan: 50 subjects in phenotype A, 101 subjects in phenotype E, and 46 subjects in phenotype M. The mean LAA scores in phenotype A, phenotype E, and phenotype M were (2.52 ± 1.73), (12.34 ± 3.59), (13.46 ± 3.93), respectively. The dyspnea severity (mMRC) and BODE index were decreased while FEV1, percent-of-predicted FEV1 (FEV1% pred), FEV1/FVC%, 6-min walking distance (6MWD), maximal inspiratory pressure (PImax), maximal expiratory pressure (PEmax) and body mass index (BMI) were increased significantly in phenotype A when compared with those in phenotype E and phenotype M (P < 0.01). There was no significant difference in the above parameters between phenotype E and phenotype M (P > 0.05). The LAA scores was positively correlated with FEV1, mMRC, COPD Assessment test (CAT), St George’s Respiratory Questionnaire (SGRQ), 6 MWD, PImax, PEmax, fat-free body mass index (FFMI), BMI and BODE index (P < 0.001). Multiple linear regression analysis suggested that BODE index, FEV1, and BMI were independent risk of LAA scores (P < 0.01). Conclusions HRCT based phenotyping of COPD patients and LAA scores may play a role in comprehensive assessment of the severity of COPD.