Background/Hypothesis: The need for a neck dissection after radiochemotherapy (RCT) for patients with unresectable cancer of the head and neck remains questionable. We evaluated our strategy to perform a neck dissection in patients with a controlled primary tumor based on the response to RCT according to regional control, survival rates, and morbidity. Study Design: The French “Groupe d'Etude des Tumeurs de la Tete et du Cou” (GETTEC) group retrospectively performed a multicenter review. One hundred and three stage III (N = 7) or IV (N = 96) patients with unresectable primary tumors and node-positive disease and no distant metastases treated between 1996 and 2002. Tumors were considered unresectable or with poor surgical curability based on advanced stage, or patients were surgically unfit for medical reasons. Results: With a median follow-up of 30 months, the complete clinical and radiological nodal response rate was 61%. Among 39% (N = 40) of patients with residual neck disease, 70% (N = 28) underwent a neck dissection, whereas the remaining 30% either underwent watchful follow-up for probable scary nodes, or were deemed unresectable or medically unfit for surgery. Half of the neck dissection specimens showed pathological evidence of viable tumor. Grade 3 to 4 complications were recorded in four patients (14%) after neck dissection. Regional control was better for complete responders. Disease-free survival and overall survival were similar between patients with a complete response in the neck and no neck dissection, and patients with a neck dissection for residual neck disease. Conclusions: The strategy to avoid a neck dissection is safe in patients with a complete response in the neck, regardless of initial nodal stage. In patients with residual neck disease, postRCT neck dissection can be performed with limited morbidity. Progress is warranted to optimize the pathological response in the nodes and to better assess ambiguous nodal responses with multi-modal imaging.