A method for diagnosing lung cancer in a subject by using a complex biomarker group is provided. The method includes steps of: (a) a computing system (1) acquiring a model M by using expression level data by individual biomarkers in the complex biomarker group and then (2) acquiring expression level data by the individual biomarkers measured from a biological specimen of the subject or their processed data Bk; and (b) the computing system determining whether lung cancer is detected in the subject by using the acquired data of the subject by referring to the model M; wherein the complex biomarker group includes CEA, HE4, ApoA2, TTR, sVCAM-1 and RANTES.