Objective: To explore the characteristics of lung segment infection in coronavirus disease 2019（COVID-19）and its clinical relevance

and to provide a certain reference for the diagnosis and treatment of COVID-19.Methods: A retrospective analysis was performed for the chest computed tomography（CT）findings and clinical data of 90 COVID-19 patients

with special attention to the number

location

and distribution of lung segments infected

and the correlation between the number of infected lung segments and clinical classification was calculated. Results: Chest CT showed no abnormalities in 8 cases

and abnormalities in 82 cases

among whom 12 had infection in a single lung segment

24 had infection in one lung

and 58 had infection in both lungs. The lung lesions were mainly patchy or spherical and appeared mostly as ground-glass opacities

with little consolidation. More than 50% of the infections occurred in the lower superior

lower lateral

and lower posterior segments. The number of infected lung segments varied significantly across different clinical types of COVID-19;the more severe the clinical type

the greater the number of infected lung segments（P < 0.05）. Conclusion: COVID-19 lesions are mostly distributed in the lower superior

lower lateral

and lower posterior segments of both lungs. There is a significant positive correlation between the number of infected lung segments and the clinical classification of COVID-19

which is helpful for clinical evaluation of the course and severity of the disease.