Establishing a biological profile of skeletal remains is a key task of forensic anthropologists. Sex estimation is essential in forensic examination, as other elements of the biological profile, such as age at death or stature, are sex dependent. Visual assessment is considered low-cost and quick, therefore it is a commonly applied method of sex estimation. The most reliable results can be obtained with the analysis of the anterior part of the pelvis, however, these skeletal elements are fragile and prone to destruction. In contrary, the more robust posterior portion of the pelvis is often recovered. Several features of the posterior pelvis have been explored in the context of sexual dimorphism. The aim of the present study was to test three previously published methods of sex assessment based on the analysis of the inferior shape of the auricular surface (Novotný, 1975), the greater sciatic notch shape (Walker, 2005 (revised)) and overall morphology, apex morphology and inflection of the auricular surface (Luna et al., 2017). The sample consisted of 194 individuals of Greek origin from a documented modern collection. Four features of the auricular surface and shape of the greater sciatic notch were examined. Logistic regression analysis was applied to produce a sex discriminatory formula. The method proposed by Luna et al. (2017) failed to produce satisfactory results with overall accuracies of 36%, 50% and 53% for overall morphology, apex morphology and inflection respectively. Slightly better results (64%) were obtained with the inferior shape morphology (Novotný, 1975). However, the highest accuracy rate of 81% was noted for the greater sciatic notch shape (Walker, 2005 (revised)). The formula produced in this study allowed correct classification of 83.2% of the sample. This study illustrates that in spite of the presence of sexual dimorphism in the posterior portion of ilium, features of the auricular surface proposed as sex indicators by Novotný (1975) and Luna et al. (2017) should not be used for sex estimation purposes in the Greek population. The formula produced in this study and the greater sciatic notch shape should only be used as additional methods in cases where neither the cranium nor the anterior portion of the pelvis is present.