Twelve thousand new cases of oral cancer occur annually in the United States. Recent laboratory developments have identified biomarkers measurable in human fluids or tissues which may help further our understanding of the process of carcinogenesis and identify persons at especially high risk of oral cancer. Among these are p53 tumor suppressor genes and heat shock proteins. Evidence from examination of dietary factors in oral cancer cases and controls, from studies of other cancers, from laboratory studies, and from some very recent chemotherapeutic trials also suggests the importance of nutritional factors in cancer etiology. This project will examine biomarkers and nutritional status in a cohort of persons with oral premalignant lesions who are monitored for changes in biomarker status or the development of cancer. Data collection activities and some laboratory assessments will be managed under contract. If the research can identify biomarkers which precede disease in a predictable way, there may be the potential to prevent oral cancer, screen for more successfully, and/or treat it at an earlier stage.