Sentinel lymph node biopsy (SNB) seems to be a promising method for staging clinically N0 neck in patients with oral squamous cell carcinoma (OSCC). In the present study, SNB was performed on 46 patients having elective neck dissection (END; six bilateral dissections) for T1-T3N0 OSCC. Sentinel lymph nodes (SLN) were first examined according to only slightly modified standard histopathologic protocol including sections at 1-2 mm intervals and H&E staining. SLN that appeared false negative (i.e. metastatic non-SLN without metastasis in a SLN) after the initial histopathologic examination were further assessed by step sectioning at 150 microm intervals and immunohistochemistry. Of the 47 neck sides with at least one SLN identified, nine contained metastasis in nine patients. After the initial histopathologic examination, SLNs were negative for malignant cells in four out of the nine metastatic neck sides. In one neck side, two metastatic SLNs were detected after the additional meticulous histopathologic work-up of the initially false negative SLNs. Therefore, in three neck sides the SLN did not contain metastasis although there was a metastasis in a non-SLN. In all these three cases with a false negative SLN, only one SLN had been identified. The sensitivity of the method (employing extensive histopathologic work-up) for detection of occult cervical metastasis was 67% (6/9 neck sides). The sensitivity of SNB for detection of occult metastasis seems to be poor in cases where only one SLN can be identified. The results of this study do not entitle us to entirely replace END by SNB in patients with OSCC.