Cathepsin B (CATB) and urokinase-type plasminogen activator (UPA) play an important part in cancer invasion and metastasis. The behavior of CATB and UPA has not been evaluated in the same experimental setting in different gastrointestinal tumors and in precancerous lesions. Serum CATB and plasma UPA levels were determined by enzyme-linked immunoadsorbent assay and their sensitivity, specificity, and diagnostic accuracy have been calculated in patients with colorectal (n=72), gastric (n=30), hepatocellular (n=28), and pancreatic cancer (n=15) as well as in gastric epithelial dysplasia (n=25), colorectal adenomas (n=30), and tumor-free control patients (n=44). Serum CATB and plasma UPA antigen concentrations were significantly higher in patients with cancer than in controls. When all tumors were considered, the sensitivity, specificity, and diagnostic accuracy of CATB (89, 86, and 89%) were higher than that of UPA (76, 70, and 74%). CATB demonstrated in all types of tumors a better diagnostic accuracy than UPA. The positive predictive values of CATB (95%) and UPA (89%) may suggest their use in the evaluation of patients with a suspicion of malignancy. CATB and UPA were significantly higher in patients with gastric epithelial dysplasia and colorectal adenomas than in controls. Antigen levels of CATB and UPA were significantly correlated in both cancers and precancerous lesions. At the time of clinical presentation, serum CATB and plasma UPA antigen levels are sensitive indicators of gastrointestinal malignancies. Determination of serum CATB and plasma UPA levels may be useful to identify patients at a higher risk for progression to cancer, who could be subjected to a more strict follow-up protocol.