OBJECTIVE: To determine the predictive value of sentinel node biopsy (SNB)-assisted neck dissection in patients with oral squamous cell carcinoma (SCC) stage T1 to 2N0M0 and to determine the incidence of subclinical metastases. STUDY DESIGN: Prospective cohort study. METHODS: Fifty-one patients with clinically N0 neck underwent SNB-assisted neck dissection. The localization of the sentinel node (SN) was determined using dynamic and planar lymphoscintigraphy and single photon emission computed tomography-computed tomography. Histopathologic examination of the harvested SN was performed using step-serial sectioning with hematoxylin-eosin (H&E) and immunohistochemistry on formalin-fixed, paraffin-embedded tissue. RESULTS: A total of 181 SNs were excised with a median of 3 (range 1-7) SNs per patient. Four percent (2 of 51) of patients with subclinical (occult) lymph node metastasis would have been identified using routine H&E staining, whereas the 18% (9 of 49) were upstaged as a result of additional histopathology when the H&E evaluation was negative. Overall, the incidence of subclinical metastases was 22% (11 of 51). CONCLUSION: In this study, SNB-assisted neck dissection proved to be technically feasible in identifying subclinical metastasis, thus accurately staging the neck with a high degree of sensitivity in patients with oral SCC T1 to 2N0M0 when additional histopathology was performed. The vast majority of patients in this study would have been spared selective neck dissection had reliance on SNB been used and selective neck dissection performed only in the case of a positive SN. Future studies should focus on determining whether SNB alone reduces patient morbidity and whether this is as equally effective in the treatment of cervical nodal metastases as compared with selective neck dissection in patients with oral SCC.