BACKGROUND Appropriate management of patients with supraglottic laryngeal carcinoma and negative findings in the neck is still controversial. A prospective and retrospective study comprised 193 patients who were treated primary surgically between 1976 and 1993. They all had clinically and ultrasound negative findings on the neck (N0). Supraglottic carcinomas usually spread regionally. Metastases develop in the jugular group, between level II-IV. The incidence of metastases has been reported to vary from 12 to 62.5%. The size and localization of the primary tumor, its histological grade, genotype of the malignant cells, immunological and other elucidated factors can all affect the incidence of regional spread. OBJECTIVE Aim of this study was to specify the incidence of occult cervical metastases; to analyze the distribution of occult metastases related to tumor localization; to specify the distribution of occult metastases related to local spread; to analyze the distribution of occult metastases according to localization in the neck. RESULTS All patients had primary surgery of primary tumor and bilateral jugular, selective neck dissection at the level II-IV with histological examination of removed lymphoid tissue. Out of 193 patients, metastatic deposits were detected in 35 (18%). Occult metastases were found in patients with carcinoma of the epilarynx in 19% (14/72) of cases, and in 17% (21/121) patients with carcinoma of the supraglottis excluding the epilarynx. This difference in frequency is not statistically significant. The incidence of occult metastases in epilaryngeal tumors did not depend on the degree of local spread. Even relatively small tumors (T1 and T2) yielded occult metastases in 33% (5/15), and 24% (6/25) of patients, respectively. In patients with T1 tumors localized at the supraglottis, excluding the epilarynx, occult metastases were not found. In the supraglottis excluding the epilarynx increased local spread was associated an increase of occult metastases. The incidence of occult metastases was directly related to the degree of the local spread of the tumor in the supraglottis excluding the epilarynx (Table 1). Occult metastases were usually ipsilateral, like the palpable ones. In medially localized tumors bilateral metastases were possible. Ipsilateral metastases were more frequent than both bilateral and contralateral ones. The possibility of contralateral and bilateral occult metastases necessitated bilateral neck dissection. Postoperative radiotherapy (60 Gy) was given to all patients with verified occult metastases. Only in two patients (1%) of the total did metastases develop subsequently, indicating the effectiveness of planned postoperative radiotherapy. CONCLUSIONS Controversies in application of jugular, selective neck dissection are present since it has been in use, because of the unclear role which regional lymph tissue play in antitumor immune response. Jugular, selective neck dissection was advocated in all patients with a primary supraglottic laryngeal carcinomas. It was suggested that selective neck dissection was needed only in advanced (T3 and T4) tumors. Selective dissection is believed to be needed only when tumor has spread into the vallecula, the base of the tongue, or the medial wall of the piriform sinus. The idea of selective neck dissection has been opposed since the protective role of the cervical lymph tissue has been stressed. Ultrasound and computerized tomography of the neck cannot detect occult metastases. Today, only removal and histological examination of the lymph tissue can determine occult metastasis. The importance of selective neck dissection is considered in diagnostic biopsy procedure by which occult metastatic spread in the neck region is established. CONCLUSIONS Due to the tendency of supraglottic carcinoma resulting in occult cervical metastases, early detection is imperative in order to apply the appropriate therapy. Occult cervical metastases are usually ipsilateral, but bilateral and contralateral may be found as well. Due to the aforementioned, it is necessary to perform bilateral jugular, selective cervical dissection of the neck level II-IV with histological evidence of removed lymph tissue. When metastases is verified histologically, postoperative radiotherapy is indicated as being efficient in hampering the development of palpable metastases. Five-year survival with no evidence disease is 86% (166/193).