In the prediction of likely disease-free and overall survival intervals in patients with squamous carcinomas of the head and neck, cervical lymph node status assumes a prime role, and patients with cervical node metastases have diminished survivals, as a group, compared with patients whose cervical nodes are reported as negative for metastatic carcinoma. Conventional means of pathologic examination of cervical node biopsy specimens include examination of a single section through each individual node identified on gross examination, a process which, of necessity, leaves a significant portion of the node unexamined by microscopy. Recently, it has become apparent that more exhaustive pathologic sampling techniques, such as examining multiple sections of each lymph node, or staining each lymph node with antibodies to keratin via immunohistochemistry, will reliably yield a greater incidence of positive cervical lymph nodes (‘micrometastases’) than do conventional pathologic techniques. This suggests that the next line of inquiry should answer this question: just because micrometastases can be detected, should they be? Does the identification of (otherwise likely to be overlooked) tiny microscopic foci of spread of tumor in regional nodes by more sophisticated techniques yield additional data of real import to the patients, or is such information of lesser value? Should a role be defined in the care of head and neck cancer patients for the use of such advanced inquiries in the structuring of therapies, then the best approach to finding such elusive micrometastases (intraoperative immunohistochemistry? immunohistochemistry using routinely fixed tissues? polymerase chain reaction?) may subsequently be established.