Background: This study evaluated the expression of p16 and ERCC1 and EGFR amplification as predictive and prognostic factors in a recently completed phase II randomized trial comparing cisplatin (100 mg/m² on d 1, 22, 43) with radiotherapy (70Gy) +/- erlotinib (150 mg/d during radiotherapy) in locally advanced SCCHN. p16 (HPV surrogate) is a known prognostic factor in oropharyngeal SCCHN, and high ERCC1 expression has been correlated with resistance to cisplatin. EGFR gene copy number may be prognostic in SCCHN and alter response to erlotinib. Methods: Pretreatment formalin-fixed, paraffin-embedded tumor tissue was available for 84/204 patients. EGFR copy number was assessed using Vysis LSI DNA probes, and immunohistochemistry was performed on Leica Bond III machines using Lab Vision ERCC-1 (8F1) and CINtec p16 antibodies. All assays were performed in accordance with manufacturer instructions; all studies were reviewed by a single Pathologist (MA) who was blinded to patient outcome. Logistic regression and Cox regression models fit a treatment arm by marker interaction and tested selected linear contrasts. Results: Efficacy analysis showed no difference in complete response rate (CRR) and progression-free survival (PFS) between treatment arms. However, PFS in both arms was better than historical comparisons, with only 46 events over a median follow-up of 23 months. Positivity in p16 was associated with greater CRR (OR 3.5, p=0.03) and longer PFS (HR=0.38; p= 0.07). The CRR effect was greater for the erlotinib arm (OR 8.1, p=0.01) than for Arm A (OR 1.5, p=0.56). The ERCC1 (+) rate was 46% (39/84).ERCC1 expression above the median was not associated with worse CRR (OR 0.69; p=0.46) or PFS (HR 0.99; p=0.98). Only 4 tumors showed EGFR amplification precluding further analysis. Conclusions: p16(+) tumors had a better outcome, similar to other recent trials, and erlotinib seemed to increase the CRR among p16(+) tumors. ERCC1 expression did not predict chemoradioresistance in this study. EGFR amplification was too rare in the tested population to assess predictive or prognostic value.