Aim The development of squamous cell carcinoma of the esophagus is a multitstep, progressive process including basal cell hyperplasia, dysplasia, carcinoma in situ and advanced esophageal squamous cell carcinoma. Chromoendoscopy with lugol solution has been found to be a valuable method to quickly screen for esophageal cancer although the meaning of unstained lesions (USL) is still unclear. The aim of this study was to evaluate the relevance of USL in a high-risk population. Patients and methods: Group 1: 87 patients with cancer of the oral cavity. Group 2: 50 patients with minimal or no alcohol or tobacco consumption. All patients underwent conventional upper gastrointestinal endoscopy, followed by chromoendoscopy with 2% lugol's solution.Tissue samples from both stained and unstained areas were obtained for subsequent histological and immunohistochemical analysis (Ki67 and p53). According to the macroscopic staining pattern, patients were divided into 3 groups (A:no USL, B: 10 or big (>5mm) USL). Results: 59 of the 84 cancer patients (70.2%; B=46.4%, C=23.8%) vs 12 of 50 patients in the control group (24%, p<.0001, all but one Group B lesions) showed USL . Among the cancer patients 4 dysplasias and 2 carcinomas could be found (4.8/2.4%). Only 1 carcinoma was detected in conventional endoscopy. Epithelial thickness and proliferation index (PI) in the control group (321.8μm±58.2/19.4%±13.7) were significantly lower than in samples from cancer patients, no matter if taken from stained (388±93μm, p<0.0001/26.3±12.1% p=0.01) or unstained lesions (477±111μm,p<0.0001/34.3±13.9%,p<0.0001). The difference between stained and USL was also highly significant(see values above,p<0.0001 for epithelial thickness,p=0.0015 for PI). p53-overexpression was more frequently observed in the cancer group (A 23.8%;B 37.1%;C 33.3%) than in the control group (4.4%,p<.05 for both). Additionally there was a clear tendency for higher values in USL as compared to stained lesions (21.4 vs 37.2%, p=0.057). Discussion: In this study we could show that hyperplasia and hyperproliferation gradually increase from normal controls via stained lesions to USL in a high risk population. Our data also provides evidence that dysplasia-associated p53 overexpression might represent an event in esophageal carcinogenesis preceding hyperproliferation and hyperplasia. Chromoendoscopy therefore seems to be a valuable and quick tool to identify potentially malignant areas in a high-risk-population. However, a homogenously stained mucosa does not rule out the existence of early premalignant alterations.