ABSTRACT Oral cavity squamous cell carcinoma (OCSCC) can be a lethal disease that is often preceded by premalignant lesions, making it is an ideal disease for screening initiatives. However, current screening protocols/tests cannot reliably differentiate between inflammatory and premalignant dysplastic lesions. Further, the histologic diagnosis of dysplasia is an imperfect predictor of malignant transformation as only ~15% of premalignant oral lesions progress to cancer. Our long-term goals are to establish molecular-based diagnostic tests for prognostication and screening that are capable of identifying high-risk patients most likely to progress to oral cancer but would greatly benefit from closer surveillance and less morbid curative intent procedures. Our central hypothesis is that premalignant lesions contain identifiable genetic mutations that can be used for reliable biopsy prognostication (tissue biopsies) and screening (saliva). We will identify dysplasia-specific mutations underlying the pathogenesis of OCSCC. We will validate the mutations identified in a retrospective case-cohort study of dysplastic oral tissues with known clinical outcomes to investigate their potential as tissue-based prognostic biomarkers. We will conduct a case-cohort study using saliva samples from five existing longitudinal population- based United States cohorts to determine whether driver somatic mutations can be identified in saliva prior to the diagnosis of oral cancer. These studies are conceptually innovative and likely to result in state-of-the-art risk stratification and screening. They would be the first to define the functional driver mutations of oral premalignancy. They would also be the first to determine if mutations in driver genes can be detected in saliva prior to oral cancer diagnosis, to define the time-course of mutation detection, and to test the predictive ability of identifying high-risk individuals with somatic mutations. They are technically innovative, as they evaluate the diagnostic accuracy of a novel non-invasive molecular salivary screening platform. This research will benefit human health by improving our ability to identify high-risk premalignant oral lesions likely to progress to cancer, thereby allowing for earlier and potentially more curative interventions with limited morbidity and mortality.