Few molecular markers have gained widespread clinical use in the practice of clinical oncology. In therapeutic clinical investigations, standardized terminology for type of study (phases 1, 2, and 3), efficacy (complete, partial response), and toxic effects (grading scales) have been accepted. In contrast, tumor marker studies have been performed haphazardly. In 1996, the American Society of Clinical Oncology convened a Tumor Marker Expert Panel to establish practice guidelines for breast and colon cancer (J Clin Oncol. 1996;14:2843-2877). Several members of the panel coauthored a proposed Tumor Marker Utility Grading System (TMUGS) that was used to evaluate each marker (Hayes, et al. J Natl Cancer Inst. 1996;88:1456-1466). The TMUGS provides a framework to define the marker designation, how it is assayed, and the proposed clinical use (risk, screening, differential diagnosis, prognosis, and monitoring). For each use, the marker only has clinical utility if it influences a clinical decision that results in an improvement in overall survival, disease-free survival, quality of life, and/or cost. A semiquantitative scale, ranging from 0 to 3 +, was generated to grade the clinical utility. Routine clinical use is recommended for tumor markers that are assigned grades of 2 + to 3 + . The assigned grade is supported by a scale of levels of evidence (LOE). Grades should only be assigned in the presence of LOE I or II data, preferably the former. Levels of evidence I and II studies provide a measure not only of the reliability (statistical significance) of the observed relative prognostic (outcome independent of therapy) or predictive (outcome related to specific therapy) values but also of the magnitude of difference between groups of patients identified by marker positivity or negativity (Hayes, et al. Br Ca Res Tr. 1998;52:304; Hayes. Eur J Cancer. 2000;36:302-306). Ultimately, these considerations will help evaluate the clinical utility of novel surrogate markers more rapidly, leading to improved patient care.