The tissue polypeptide antigen (TPA) is a protein produced and released by proliferating cells that possesses several characteristics for an ideal tumor marker. Our purpose was to define the clinical yield of TPA in the follow-up of patients with lung cancer (LC). Three hundred and forty-one new LC patients underwent an extensive pre-treatment staging evaluation (UICC 1987 classification) and a TPA serum measurement. We restaged them at regular times by: 1, clinical history and physical examination, routine lab tests, chest X-rays, and any other examination as suggested by the prior baseline evaluation, and 2, the serum level of TPA. We evaluated a total of 1513 assays (including 1172 posttreatment measurements). Individual values of TPA correlated significantly with treatment response and disease status. Patients with small-cell lung cancer showed the lowest correlation indexes between clinical parameters and the marker. Each objective response to treatment or disease progression was almost always associated to consistent changes of TPA (P < 0.0001, by the Wilcoxon's test). A 50% reduction under the prior TPA value was 30% sensitive, 90% specific, and 88% accurate in the diagnosis of response to treatment. The same percent reduction was 18% sensitive, 92% specific, and 88% accurate in predicting a future response. A 100% increase over the prior level of TPA permitted to recognize tumor relapses with sensitivity, specificity, and accuracy of 30%, 93%, and 80% (diagnosis of progression), and 18%, 92%, and 78% (prediction of progression). Similar diagnostic yields were observed using progressively increasing or progressively decreasing changes of the marker level. In lung cancer, the diagnosis (and even the anticipation) of disease status is often possible using appropriate threshold value of TPA.