We retrospectively selected 22 cases in which patients with a biopsy-proven diagnosis of cervical intraepithelial neoplasia grade 3 underwent cervical conization for frozen section (FS) evaluation followed by hysterectomy at the University of California Irvine Medical Center, Orange, during the August 1995 to September 9, 2001. All slides from FS and permanent section (PS) and hysterectomy specimens were reviewed. FS diagnoses were compared with those of previous biopsies, PS, and hysterectomy specimens. The PS correlated with FS in all cases but 1. Appropriate surgery was performed for all patients based on FS diagnosis. The McNemar test was used to compare the results of FS and PS, with a 2-sided P value of 1.0 and a c coefficient of 0.7755 with a 95% confidence level, indicating that the 2 groups were not significantly different. FS evaluation of cervical conization is as efficacious and accurate as evaluation of regular specimens in providing information for the appropriateness of same-day surgery. We recommend that entire tissue be submitted for FS to avoid sampling errors and to increase diagnostic accuracy.