The purpose of this study is to assess the accuracy and limitations of the outpatient endometrial biopsy in detection of endometrial carcinoma and atypical hyperplasia. This was done by reviewing all the endometrial material from the 891 women at this clinic whose first endometrial sample in 1974 was by the biopsy technique. The results of this material were correlated with subsequently available additional uterine samples obtained over a three year period by biopsy, dilatation and curettage, or hysterectomy. The detection rate by biopsy for the 21 cases of endometrial carcinoma proven by hysterectomy was 81%. There was a 5% false negative rate and no false positives. For the 25 cases of atypical hyperplasia proven by hysterectomy, the detection rate by biopsy was 68%. There was lack of hysterectomy confirmation in 33% and 24% false negatives. Insufficient material was obtained initially in 9% of the 891 women. Five percent of this group were found at hysterectomy to have either endometrial carcinoma or atypical hyperplasia. With an understanding of its limitations, the endometrial biopsy, although somewhat less accurate than the D & C, is considered to be a useful diagnostic tool for detection of these lesions.