Abstract Objective: To assess the independent contribution of transvaginal ultrasound in identifying women at risk for endometrial disorders, and determine whether a cutoff value identifies women who need endometrial histologic assessment. Methods: Postmenopausal women with breast cancer who were receiving tamoxifen, with ultrasonographic endometrial thickness greater than 4 mm or vaginal bleeding, had hysteroscopy with selective endometrial biopsies. Endometrial thickness, duration of tamoxifen therapy, and endometrial histology were studied. Parametric and nonparametric tests and logistic regression and receiver operating characteristic curves were used for statistical analysis. Results: The study population consisted of 163 women, 46 with vaginal bleeding. The proportion of women with abnormal histologic findings was greater among those with endometrial thicknesses exceeding 9 mm compared with those with endometrial thicknesses 9 mm or less (60% versus 6.1%, P < .001) and among women who received tamoxifen for more than 27 months than those who received it for less time (46% versus 16%, P < .005). Logistic regression showed that endometrial thickness greater than 9 mm and vaginal bleeding were independent predictors of abnormal findings at hysteroscopy. Conclusion: In women taking tamoxifen, sonographic endometrial thickness exceeding 9 mm and the presence of vaginal bleeding are independent predictors of endometrial disease. If either exists, hysteroscopy and biopsy should be done.