To evaluate the correlation between the ability to predict the attachment site of sinonasal inverted papilloma by computed tomography and the long-term surgical outcome. Retrospective cohort study. Five tertiary medical centers. Study patients underwent attachment-oriented resection of inverted papilloma. The primary outcome was tumor recurrence. Among 195 patients eligible for the study, focal hyperostosis was recognized on computed tomography in 65% (n = 127), in 71% of primary cases (n = 101), and in 50% of revision procedures (n = 26). There was a trend for a higher incidence of squamous cell carcinoma among the patients without detectable hyperostosis (P = .051). Location of hyperostosis coincided with the actual tumor attachment site in 114 patients (90%). Discordance between these parameters did not differ significantly (P = .463) between 11 primary and 2 revision cases. The overall rate of recurrence was 9.7% (n = 19), with a mean time to recurrence of 20 months (range, 7-96 months). The rate of recurrence did not correlate with any of the following: tumor stage, surgical approach, presence of squamous cell carcinoma, whether the surgery was primary or revision, and the presence or location of focal hyperostosis on computed tomography. Inverted papilloma recurred significantly more often (38.5%) when the intraoperative findings of the tumor attachment site did not match the location of hyperostosis observed on computed tomography (odds ratio, 6.5; 95% CI, 1.78-23.66). Detectability of focal hyperostosis on preoperative computed tomography does not affect the long-term outcome of inverted papilloma resection.