The present review regarding atypical endometrial hyperplasia (AEH) focused on the main debated factors regarding this challenging clinical condition: (i) predictive variables of occult endometrial cancer (EC); (ii) the rate of EC underestimation according to different endometrial sampling methods; and (iii) the appropriateness of lymph node status assessment. When cancer is detected, approximately 90% of cases include low-risk EC, although intermediate/high-risk cases have been found in 10-13% of women with cancer. Older age, diabetes, high BMI, and increased endometrial thickness are the most recurrent factors in women with EC. However, the predictive power of these independent variables measured on internal validation sets showed disappointing results. Relative to endometrial sampling methods, hysteroscopic endometrial resection (Hys-res) provided the lowest EC underestimation, ranging between 6 and 11%. Further studies, including larger sample sizes of women undergoing Hys-res, are needed to confirm these findings. These data are urgently needed, especially for female candidates for conservative treatment. Finally, the evaluation of lymph node status measured on 660 of over 20,000 women showed a lymph node positivity of 2.3%. Although there has been an increase in the use of this procedure in AEH in recent years, the present data cannot recommend this option in AEH based on a cost/risk/benefit ratio.