OBJECTIVE The porous hydroxyapatite (HA) tubiform prosthesis was employed for reconstruction of large circumferential defect of the larynx and cervical trachea. The biocompatibility, bioactivity and biofunctionality of the HA prosthesis were evaluated, and the feasibility for laryngotracheal reconstruction was discussed. METHODS Twelve healthy canines were used to establish the experimental models, a large portion of cricoid cartilage and upper ten-ring segment of cervical trachea were resected and substituted with a corresponding 5 cm HA prosthesis by end-to-end anastomosis. Six months later, the eight survival canines were sacrificed and the HA prostheses with surrounding tissues were removed. They were observed in decalcified sections by optical microscopy and scanning electron microscopy. RESULTS Two canines died in the immediate postoperative period with unknown reason and another two deaths were attributed to obstruction caused by the complete dislocation of the HA prostheses owing to rupture of the sutures within three weeks after operation. Eight canines survived up to six months. The implanted HA prostheses were tolerated in all cases without any rejection and dislocation or shift. An excellent airway was obtained and no signs of dyspnea and suffocation were found though there were hypergranulation and scar formation occurred at the site of anastomosis. Morphologic examination revealed that collagen fibers, new vessels and plenty of cells penetrated deeply into the pores of HA, and occupying the outer two third of HA wall. HA prostheses were surrounded by connective tissues and anchored firmly to the neighboring tissues, including the ends of the cricoid and tracheal cartilage by ingrowths of cartilaginous tissue into the macropores of the HA. However, the luminal surface of HA prosthesis was not covered at all section levels by respiratory mucosa. CONCLUSIONS The implantation of the porous HA tubiform prosthesis can maintain the normal respiratory function of the experimental canines, but the proliferating granulation and scar formation at the anastomotic site are questions still remained to be solved. To cover the inner surface of HA with the epithelial mucosa and then reduce the morbidity caused by scar and hypergranulation, some forms of allografts such as pedicled flaps and jejunal autografts will be deserved.