It is a well-known fact that the clinical growth of laryngeal cancer varies according to the site of its origin. A number of macro and microscopy studies have attempted to explain the various routes it follows in spreading. The anatomical features of the anterior commissure, the conus elasticus and posterior commissure condition the direction and extent of tumour spreading as do the blood supply in the glottis and the distribution of mucous glands in the glottis and subglottis. Of the various regions of the larynx, the anterior commissure has provided perhaps the greatest challenge to investigators in that its boundaries as well as its morphology are still not clear. The possible spreading of glottic cancers exceeding these anatomic structures is very difficult to evaluate using two dimensional picture given by laryngoscopy as well as the images provided by TC and NMR which justifies a different behavior of same stage. Therefore these elements, in addition to the absence of a universally accepted clinical definition of the boundaries of AC, justify the diagnostic, pathologic, and therapeutic problems linked to AC carcinoma. Eight normal adult larynges were studied by microdissection and serial section after fixation in 10 percent formalin followed by decalcification. These specimens were dissected according to the stages of "evisceratio laryngis" performed in our Department for the treatment of T1a, T1b tumors. Our observations confirm that the island located at the anterior insertion of the thyroarytenoid muscle, easy reached during subperichondral dissection, cannot be identified with the so-called tendon described by Broyles. The specimens shown circumstantiate the fact that this fibrous-cartilagineous island act as a barrier and that when the neoplasia does start to spread into this segment, this fibrous area forces it to spread mainly along the surface. Our observations were confirmed by histologic examination of 6 surgical specimens after "evisceratio laryngis" performed on T1a, T1b tumors. Clinical positive results in subjects treated employing "evisceratio laryngis" appear to further confirm of our interpretation to these morphological observations.