Currently available oral cancer screening adjuncts have not enhanced clinical screening methods because of high false positives and negatives, highlighting the need for a molecularly specific technique for accurate screening of suspicious oral lesions. The purpose of this study was to evaluate the in vivo screening accuracy of an oral lesion identification system that evaluates aberrant glycosylation patterns using a fluorescently labeled lectin (wheat germ agglutinin and fluorescein isothiocyanate [WGA-FITC]).The authors designed and implemented a prospective cohort study at 3 institutions composed of patients with and without suspicious oral lesions. Oral cavities were screened by clinical examination and with the oral lesion identification system according to a stepwise procedure that included the topical application and fluorescence visualization of a fluorescent nuclear stain and WGA-FITC. Tissue samples were obtained from all enrolled patients for histopathological diagnosis and were used to calculate sensitivity and specificity metrics (primary outcome variable) irrespective of the oral lesion identification system result.The sample was composed of 97 patients; 86 had 100 clinically suspicious lesions and 11 without such lesions were included as a control group. Use of the oral lesion identification system resulted in 100, 100, and 74% sensitivity for cancer, high-grade dysplasia, and low-grade dysplasia, respectively, and a specificity of 80%. Clinical diagnosis yielded similar sensitivity values of 84, 100, and 88% for cancer, high-grade dysplasia, and low-grade dysplasia, respectively, and a specificity of 76%. Use of the oral lesion identification system enhanced the visualization of lesion dimensionality and borders.The results of this study suggest the oral lesion identification system was a beneficial adjunct to standard clinical examination, because the system provided sensitivity and specificity values similar to or greater than clinical diagnosis.