Routine dilatation and curettage (D&C) in all patients with postmenopausal bleeding (PMB) is debatable, as 70%-80% will eventually receive benign diagnoses. Endometrial thickness (ET) measurements by transvaginal ultrasonography (TVUS) are used with high sensitivity to detect patients who would benefit from D&C, yet they suffer from low specificity that fails to reduce undue invasive procedures. The aim of this study was to define optimal cutoffs for ET in diagnosing endometrial pathologies in PMB and to assess a possible complementary role for Doppler ultrasonography. The study population consisted of 97 women with PMB; 39, 22 and 36 of whom had endometrial cancer (EC), benign endometrial pathologies and normal endometrial findings, respectively, defined by D&C performed after TVUS, which was used to measure uterine dimensions and ET, together with pulsatility and resistance indices (PI and RI, respectively) of the uterine arteries. Receiver operating characteristics curves revealed ET to be the most valuable parameter to prognosticate both EC and any endometrial pathology (sensitivities of 90% and 89%, and specificities of 79% and 94% with optimal cutoffs of 9.6 and 7.7 mm, respectively). Binary logistic regression revealed uterine artery RI to be the only independent variable that could be used together with ET, which increased the sensitivity of ET to 97% and 93% for EC and any endometrial pathology, but caused its specificity to regress to 58% and 53%, respectively. Same levels of sensitivity, yet better levels of specificity of 60% and 89%, respectively, were attained by using a cutoff of 6.3 mm for ET alone. Assessing uterine artery Doppler indices has no complementary role for measuring ET in evaluating PMB.