The search for and the application of available noninvasive methods for early diagnosis of oral mucosa (OM) neoplasia is a clinically significant problem. The aim of this study was to evaluate the effectiveness of the original score-based algorithm for assessing clinical data generated by a conventional and an autofluorescencebased examination in diagnosing OM cancer and assessing indications for a biopsy. We analyzed 134 medical histories and pathology reports of patients with oral neoplasia. The patients were assigned to 2 groups: the control group included 63 patients who underwent a standard visual and tactile examination with history taking and then were referred for an incisional biopsy followed by a histopathological examination of the specimens. In the main group consisting of 71 patients, a standard visual and tactile examination was complemented by an autofluorescence-based examination and the original score-based algorithm with the original index of required histopathological verification (RHV) were used to assess indications for a biopsy. In both groups, the most commonly affected site was the tongue (72.4%). The histopathological examination revealed that 28 patients from the main group and 14 patients from the control group had OM cancer (р = 0.051). Histologically, early-stage cancer was diagnosed in 17 patients from the main group and in 4 patients from the control group (р = 0.004). The proposed algorithm allowed us to effectively (in 90% of cases) diagnose precancer and cancer and avoid unnecessary biopsies.