Development of head and neck squamous cell carcinoma (HNSCC) is a multistep process and in many cases involves a phenomenon coined 'field cancerization'. In order to identify changes in protein expression occurring at different stages of tumorigenesis and field cancerization, we analysed 113 HNSCCs and 73 healthy, 99 tumor-distant and 18 tumor-adjacent squamous mucosae by SELDI-TOF-MS on IMAC30 ProteinChip Arrays. Forty-eight protein peaks were differentially expressed between healthy mucosa and HNSCC. Calgizarrin (S100A11), the Cystein proteinase inhibitor Cystatin A, Acyl-CoA-binding protein, Stratifin (14-3-3 sigma), Histone H4, alpha- and beta-Hemoglobin, a C-terminal fragment of beta-hemoglobin and the alpha-defensins 1-3 were identified by mass spectrometry. The alpha-defensins showed various alterations in expression as validated by immunohistochemistry (IHC). Supervised prediction analysis revealed excellent classification of healthy mucosa (94.5% correctly classified) and tumor samples (92.9% correctly classified). Application of this classifier to the tumor-adjacent and tumor-distant mucosa samples disclosed dramatic changes: only 59.6% of the tumor-distant biopsies were classified as normal, 27.3% were predicted as aberrant or HNSCC. Strikingly, 72% of the tumor-adjacent mucosae were predicted as aberrant. These data provide evidence for the existence of genetically altered fields with inconspicuous histology. Comparison of the protein profiles in the tumor-distant-samples with clinical outcome of 32 patients revealed a significant association between aberrant profiles with tumor relapse events (P=0.018; Fisher's exact test, two-tailed). We conclude that proteomic profiling in conjunction with protein identification greatly outperforms histopathological diagnosis and may have significant predictive power for clinical outcome and personalized risk assessment.