OBJECTIVE/HYPOTHESIS: To assess the oncologic efficacy and functional outcome of selective postchemoradiotherapy neck dissection for stage IV head and neck squamous cell carcinoma. METHODS: Retrospective review of patients with N2-3 cervical metastases at presentation who underwent planned neck dissection after complete biopsy-proven clearance of primary site mucosal disease with chemoradiotherapy between 2000 and 2006. RESULTS: There were 31 males and 10 females. The average age at presentation was 57 +/- 9 years. The oropharynx was the most common primary site (n = 23; 56%). Forty-nine hemineck dissections were performed, including six bilateral and two revision procedures. Sixteen (39%) patients had residual viable postchemoradiotherapy neck disease. Patient weight did not deteriorate after neck dissection (P > .4). Two patients had persistently worsened postoperative swallowing. Ten patients required shoulder physiotherapy, of whom eight were treated with conservative measures. Five-year hemineck disease control and disease-specific survival rates were 92% and 64%, respectively. Presence of viable postchemoradiotherapy neck disease was the only independent predictor of regional control (P < .001; hazard ratio 0.00; 0.00-0.40) and disease-specific survival (P < .02; hazard ratio 0.23; 0.04-0.55). Surgery was twice more likely to confer therapeutic benefit than to cause a significant, albeit in most cases, transitory, complication. CONCLUSIONS: Neck dissection is a safe and effective procedure and a necessary component of the multimodality management of all head and neck cancer patients with N2-3 disease. It should be performed soon after satisfactory demonstration of primary site disease clearance. Universal deployment of radical surgery appears unnecessary and should, when possible, be abandoned in favor of more selective procedures to lessen morbidity.