This study was designed to analyze certain clinicopathological features and the profile of p53, Ki-67, estrogen (ER), and progesterone (PR) receptor expression of clear cell carcinoma of the endometrium and to determine whether the pathogenesis of clear cell carcinoma can be accommodated by a dualistic model of endometrial carcinogenesis. In this model, endometrioid carcinoma develops from endometrial hyperplasia under unopposed estrogenic stimulation, and serous carcinoma develops in atrophic endometrium from a putative precursor lesion designated endometrial intraepithelial carcinoma (EIC). Twenty-one clear cell carcinomas of the endometrium were analyzed and compared with 77 endometrioid carcinomas of all grades and 30 serous carcinomas. Clear cell carcinomas showed a distinctive immunoprofile characterized by immunonegativity for ER and PR, low immunoreactivity for p53, and a high Ki-67 proliferation index. ER, PR, and Ki-67 expression were similar to serous carcinoma, but p53 expression was significantly lower in clear cell carcinoma (P < .05). ER and PR expression were significantly lower, and the Ki-67 proliferation index was significantly higher in clear cell carcinoma compared with endometrioid carcinomas (P < .05). p53 expression tended to be higher in clear cell carcinoma compared with endometrioid carcinoma, but the difference was not statistically significant. In contrast to endometrioid carcinoma, clear cell carcinoma was rarely associated with endometrial hyperplasia and serous carcinoma was not. Subdividing clear cell carcinoma morphologically into one that resembled serous carcinoma (clear cell carcinoma with serous features) and another that did not (typical clear cell carcinoma) showed that clear cell carcinoma with serous features had a higher Ki-67 proliferation index than typical clear cell carcinoma, although expression of ER, PR, and p53 were similar. Clear cell carcinoma with serous features was associated with EIC in 50% and was not associated with endometrial hyperplasia. In contrast, typical clear cell carcinoma was associated with endometrial hyperplasia in 40% and was not associated with EIC. In summary, this study provides evidence that clear cell carcinoma of the endometrium, like serous carcinoma, is estrogen independent and shows a high Ki-67 proliferation index. In contrast to serous carcinoma, strong p53 expression occurred less frequently in clear cell carcinoma and predominantly in clear cell carcinoma with serous features. The findings suggest that the molecular events that underlie the development of clear cell carcinoma differ from those of endometrioid and serous carcinoma.