Thus, many interesting and promising techniques are now available for the study of potential prognostic factors in esophageal carcinoma, and the results may supplement the information that is obtained by routine histopathologic analysis. It may become possible to define subgroups of patients with lower risk tumors, for whom more aggressive therapy is indicated, especially within the group of early-stage lesions. It is clear, however, that much more data is necessary to define the role each study should play in predicting tumor behavior. While the relevance of the newer techniques is still being established, it is important that all of the data from these studies be incorporated into ongoing research studies, since in this era of cost-conscious medicine it may soon become impossible outside of a research setting for pathologists to perform large batteries of special studies to define parameters of unproven prognostic value.