Despite the high resolution of ultrasound or magnetic resonance imaging and modern bioptic approaches, diagnostic error of preoperative staging is around 20%. The preoperative staging is focused to differentiate low risk (< 50% myometrial invasion, histological grade 1 and 2, endometrioid cancer, no cervical invasion) or high risk endometrial cancer (all others). Preoperative biopsy tends to underestimate the tumour histotype and grading in 20% of cases. Therefore, the sonomorphological and Doppler pattern have been identified that can preoperatively alert us to the presence of low or high risk endometrial cancer. With discrepancy between preoperative ultrasound tumour characteristics and results of endometrial biopsy a supplementary intraoperative frozen section of uterus is recommended. Ultrasound examination performed by an experienced sonographer is accurate in most cases but tends to overestimate myometrial invasion and underestimate cervical stromal invasion. Based on the identification of factors that significantly affected the accuracy of ultrasound, it was recommended to restrict evaluation to sonomorphological tumour parameters within the preoperative tumour staging. This is in response to the tendency of ultrasound experts with knowledge of the clinical data of patients, and prognostic parameters of the disease who have encountered adverse tumour characteristics through ultrasound (e.g. inhomogeneous, hypo- or isoechogenous tumor with high tumour perfusion) to 'intuitively overestimate the stage of the disease and conversely. The attempts to restrict the assessment of tumour invasion to those parameters that are easily and objectively evaluated has not so far proved effective. A promising objective parameter seems to be the minimum distance between the tumour and uterine serosa. When this parameter was included in the new objective model, the high-risk endometrial cancer was predicted with an accuracy similar to subjective assessment of tumour invasion by an expert. The preoperative work-up for high-risk endometrial cancer prediction was designed and externally validated in order to triage the patients for adequate staging surgery. This approach was based on the combination of the results of preoperative endometrial biopsy and transvaginal ultrasound and reached the similar accuracy as a more complicated approach using a combination of magnetic resonance imaging and hysteroscopic-directed biopsies. Both approaches can identify eight of 10 women with high-risk of lymph node metastases and spare eight of 10 low-risk women extended surgery.