This study aimed to evaluate the diagnostic accuracy of the sentinel lymph node (SLN) mapping algorithm in high-risk endometrial cancer patients.Two hundred forty-four patients with non-endometrioid histology, grade 3 endometrioid tumors and/or tumors with deep myometrial invasion were enrolled in this retrospective, multicentric study. After removal of SLNs, all patients underwent pelvic ± paraaortic lymphadenectomy. Operations were performed via laparotomy, laparoscopy or robotic surgery. Indocyanine green (ICG) and methylene blue (MB) were used as tracers. SLN detection rate, sensitivity, negative predictive value (NPV) and false-negative rate (FNR) were calculated.Surgeries were performed via laparotomy in 132 (54.1%) patients and 152 (62.3%) underwent both bilateral pelvic and paraaortic lymphadenectomy. At least 1 SLN was detected in 222 (91%) patients. Fifty-five (22.5%) patients had lymphatic metastasis and 45 patients had at least 1 metastatic SLN. Lymphatic metastases were detected by side-specific lymphadenectomy in 8 patients and 2 patients had isolated paraaortic metastasis. Overall sensitivity, NPV and FNR of SLN biopsy were 81.8%, 95% and 18.2%, respectively. By applying SLN algorithm steps, sensitivity and NPV improved to 96.4% and 98.9%, respectively. For grade 3 tumors, sensitivity, NPV and FNR of the SLN algorithm were 97.1%, 98.9% and 2.9%.SLN algorithm had high diagnostic accuracy in high-risk endometrial cancer. All pelvic metastases were detected by the SLN algorithm and the isolated paraaortic metastasis rate was ignorable. But long-term survival studies are necessary before this approach becomes standard of care.