This study aimed to compare endometrial cavity length measurements obtained by preoperative three-dimensional (3D) reconstruction of coronal images on pelvic ultrasound with intraoperative endometrial cavity measurements obtained in advance of Minerva endometrial ablation. This was a retrospective cohort study of 41 individuals who underwent a Minerva endometrial ablation between August 1, 2018 and March 15, 2022 at a single academic medical center. Patients were excluded if they had an in-clinic ablation or no ultrasound with 3D uterine volume within 180 days before surgery. Physician sonologists measured the endometrial cavity length using 3D coronal reconstruction of the cavity. Two measurements were obtained by separate physicians who were blinded to intraoperative values. Intraoperative endometrial cavity lengths were obtained from operative reports. A Bland-Altman plot was used to evaluate the agreement of intraoperative and ultrasound measurements. The mean intraoperative endometrial cavity length (50.7 ± 7.8 mm) was greater than the mean endometrial cavity length by 3D coronal reconstruction of pelvic ultrasound (36.1 ± 6.2 mm, P < 0.0001). The average difference between intraoperative and ultrasound measurements of cavity length was 14.6 ± 9.0 mm. The agreement between measurements was poor (Lin's concordance correlation coefficient 0.06). Using a Bland-Altman plot, the limits of agreement (-3.1 to 32 mm) exceeded the a priori acceptable limits of agreement (-10 to 10 mm). Our findings suggest that preoperative coronal endometrial cavity length measurements by ultrasound are not a valid substitute for intraoperative measurements before Minerva ablation.