Although production of immunosuppressive factor(s) by esophageal squamous cancer has been demonstrated, systemic immunosuppression occurs late. Whether local immunosuppression by tumor-derived factors occurs in vivo as a potential mechanism of escape from immune surveillance is unknown. We found that lymphocytes from nodes draining distal esophageal squamous tumors in 23 consecutive patients had depressed proliferative and cytotoxic responsiveness relative to both lymphocytes from a reference node outside the field of drainage and matched PBL from the same patient. In a subset of patients in which more than one tumor-draining node was examined, a radial or zonal immunosuppression relative to the primary tumor was evident. The findings were unrelated to surgery or anatomic location because all but 2 of 26 control patients with esophagogastric adenocarcinoma had normal or enhanced lymphocyte responsiveness in the tumor-draining node. The absence of overt or even micrometastatic nodal disease, as determined by immunostaining for cytokeratin expression, coupled with the long-term survival of several of the patients, strongly suggests that the immunosuppressive effect is due to mechanisms other than metastases, and may be a premetastatic occurrence. We conclude that regional immunesuppression does exist in patients with esophageal squamous cancer when systemic immunity is still well preserved. The local immune suppression inhibits the generation of lymphokine-activated killer (LAK) cells and may be an impediment to potential immunotherapeutic strategies.