One hundred six patients with histologically proven bronchogenic carcinoma were tested for carcinoembryonic antigen (CEA), tissue polypeptide antigen (TPA), and carbohydrate antigenic determinant 19-9 (CA19-9). A total of 349 CEAs, 350 TPAs, and 317 CA19-9s were measured. In addition, sera were assayed from 57 patients with pulmonary benign diseases and their CEA, TPA, and CA19-9 levels were used as negative controls for specificity and accuracy. One hundred twenty healthy subjects provided our normal CA19-9 reference value. Sensitivity, specificity, and accuracy were obtained for CEA, TPA, and CA19-9, respectively. Significant intermarker correlations were found both at diagnosis and during follow-up, CEA and CA19-9 being the most closely related substances. The percentage of patients with elevated levels of TPA increased significantly according to tumor load. Individual values of TPA related significantly to the stage of disease. Concentrations of CEA, TPA, and CA19-9 varied significantly during the course of the illness in relation to treatment response; however, TPA showed the closest relationship to the clinical status assessments of the follow-up period. Abnormal pretreatment levels of TPA were significantly associated with a poor outcome. Biomarker combinations were clinically evaluated by calculating the mean of the percentage of the reference value for each combined marker. Using this method, any association of TPA with CEA and/or CA19-9 revealed neither a greater diagnostic accuracy nor a more reliable predictive capacity for the above clinical variables than TPA evaluated on its own. The authors believe that a single TPA assay should be added to the initial and subsequent clinical assessments of patients with bronchogenic carcinoma.