Patients with superior dehiscence (SCD) syndrome present with vertigo and oscillopsia evoked by loud sounds and changes in middle ear or intracranial pressure. The first objective of this retrospective cohort study is to demonstrate that thin-section computed tomography (CT) scans reformatted in the plane of the superior semicircular canal (SSC) overestimate this anomaly compared to pathologic studies. The second objective of this study is to re-evaluate the positive predictive value of temporal bone scanning. All temporal bone CT scans with 0.55-mm collimation and reconstruction in the SSC plane performed over a 1-year period were analysed at a tertiary referral centre. CT-positive cases had their clinical data reviewed and patients were re-examined, if available. A total of 581 temporal bone CT-scans were analysed. A dehiscent-appearing superior canal was seen in 4.0% of studies while published pathologic studies report that only 0.5% of temporal bones SSCs have a dehiscence (P < 0.001). Of the 21 patients with positive temporal bone CTs, only 1 presented with sufficient clinical dues to identify the syndrome. Three additional patients did not have symptoms consistent with the diagnosis, but had surgery for a dehiscence of the tegmen mastoideum. When our findings are added to published data, the positive predictive value of temporal bone CT-scanning drops from 93 to 57%. The prevalence of dehiscent-appearing superior canal on thin-section temporal bone scanning with reformation in the SSC plane is much higher than anticipated by pathologic studies. Even with 0.55 mm-collimated helical CT and reformation in the SSC plane, the risk of overdiagnosis is present.