Purpose/Objective: Postoperative irradiation (XRT) following surgical resection is the often recommended treatment for resected advanced primary and resected recurrent squamous cell carcinoma (SCC) of the head and neck. Despite common use, there is little literature comparatively evaluating the results of such a management approach between those patients irradiated postoperatively for primary versus recurrent disease. This study evaluates the efficacy of postoperative XRT in patients with recurreni SCC of the head and neck and compares their outcome to patients with primary SCC of the head and neck treated with postoperative XRT over the same time period. Materials & Methods: Between 1981 and 1993, 174 patients with SCC of the head and neck, 143 with primary and 31 with recurrent disease, were treated with standard postoperative XRT. All had complete resection of gross disease, no distant metastasis, or prior XRT. All patients were treated with non-split course XRT at 1.8-2.0 Gy per fraction. Median total postoperative XRT dose was 66 Gy (range = 50. 72Gy). Median follow-up is 68 months (range = 22-159 months). Several potential prognostic factors for disease-specific survival (DSS) and local-regional control (LRC) were analyzed and included tumor type (primary vs. recurren& site and stage of original tumor, status of surgical margins, disease-free interval, site and extent of recurrence, and dose. Results: For all 174 patients, the .5-year DSS and LRC rates were 50% and 65%, respectively. On univariate analysis, both tumor type and original tumor site significantly influenced DSS and LRC. Patients treated forprimary disease had 5.year DSS and LRC rates of 54% and 69%, respectively, as compared to 32% and 46%, respectively for patients treated for recurrem disease (P=O.O4 and 0.03, respectively). Patients with a non-oral cavity original tumor had 5-year DSS and LRC rates of 55% and 73%, respectively, as compared to 39% and 44%, respectively, for patients with an oral cavity original tumor ( P=O.O5 and 0.002, respectively). On multivariate analysis, only primary tumor site significantly influenced DSS (P=O.O4) and only tumor type significantly influenced LRC (P=O.O03). Among the 143 patients treated for primary disease, no factors analyzed significantly influenced DSS, while on multivariate analysis, only surgical margin status significantly influenced LRC. Patients treated for primary disease with negative margins had a 5-year LRC rate of 78% as compared to 61% for those with positive /close margins (P&03). Among the 3 1 patients treated for recurrent disease. the factor which significantly influenced both DSS and LRC on multivariate analysis was site of recurrence. Patients whose recurrence was limited to previously undissected tissue had 5-year DSS and LRC rates of 60% and 74%, respectively, as compared to 19% and 29%, respectively, for patients whose recurrence was within previously dissected tissue (P=O.OOl and 0.008, respectively). Conclusion: Patients with recurrent SCC of the head and neck do poorly when treated with standard postoperative XRT following surgical resection, especially if the recurrence is within previously dissected tissue, as compared to patients treated forprimaiy disease. This may be related to the negative effects of hypoxia on the efficacy of XRT or an intrinsically resistant tumor clone. This patient group should be targeted for alternative local treatment strategies, including dose escalation and/or altered fractionation.