Abstract: Objective To investigate the factors of missed diagnosis of lung cancer on chest X-ray radiography (CXR). Methods A total of 460 patients with lung cancer (467 lesions) which confirmed by surgical pathology were consecutively collected from June to December 2014. All patients had CXR and CT examination before operation. Patients were categorized into two groups: lung cancer negative group and positive group. Lesion size, density, location, imaging classification, pathological type, obstructive change were compared between two groups. Then multivariate regression analysis was performed to detect the independent risk factors of missed diagnosis of CXR. Results In 467 lesions, total missed diagnosis rate of lung cancer on CXR was 25.9%. It was 100% in lesions smaller than 10 mm, and 31.1% in lesions between 10 mm and 20 mm, whereas it was 4.7% in lesions greater than 20 mm. Univariate analysis showed lung cancer negative group and positive group had significant differences in lesion size, density, imaging classification, pathologic types and obstructive change. Multivariate regression analysis showed that lesion size, density and imaging classification were independent risk factors of missed diagnosis (P< 0.05). Conclusion Lesions size, density and imaging classification are the main risk factors of missed diagnosis of lung cancer on chest X-ray radiography. Chest X-ray radiography is not suggested for detection of early lung cancer but low-dose CT.