PURPOSE: To assess the usefulness of three-dimensional (3D) spiral CT imaging in patients with upper airway stenosis. MATERIALS AND METHODS: We performed 3D spiral CT imagings in ten patients in whom upper airway stenosis was clinically suspected. Eight of these patients had upper airway stenosis caused by intubation or tracheostomy(n=6), tuberculosis (n=1), or extrinsic compression by a thyroid mass (n=1). Spiral CT scanning(30-second continuous exposure and 90-mm length) was performed with a table speed of 3mm/sec and a section thickness of 3mm.The selected starting point was the epiglottis. The resulting data were reformatted by multiplanar reformation(MPR) and shaded surface display (SSD) with peeling after reconstruction of 2mm interval. In the evaluation oflocation and extent of stenosis, we compared fidings of 3D imaging with those of baseline axial images (n=10), endoscopy (n=9) and operation (n=4). RESULTS: The locations of stenosis in eight patients were as follows: tracheostoma (n=4), subglottic region (n=3), and larynx (n=1). In all eight patients, 3D imaging demonstrated the location and extent of stenosis, which exactly correlated with endoscopic and operative findings. In one patient, however, another stenotic area in the tracheal bifurcation was not discovered because this lesion was not includedin the field of CT scan. In two patients, the diagnosis on 3D images of 'no stenosis' was comfirmed by clinical findings or operation. No differences in diagnostic accuracy were noted between axial images, MPR, and SSD when evalvating the location and extent of stenosis ; vertical extent was shown more easily by 3D imaging than by axialimages, however. CONCLUSION: 3D imaging with spiral CT may be an useful adjunctive method in the evaluation of upper airway stenosis with variable causes.