Primary ovarian malignant mixed mesodermal tumors are uncommon. There exist few data in the literature on the significance of the sarcomatous component (SC) in these tumors. Here we investigated this aspect in 47 such tumors, with particular interest in whether the presence of SC outside the ovary confers a worse prognosis. We correlated various features of the SC (homologous vs. heterologous, type of heterologous SC, extent/percentage, mitotic count, necrosis, whether or not SC is present outside the ovary) with disease-specific survival (DSS) using the Kaplan-Meier method and log-rank test. We also correlated other clinicopathologic parameters with DSS: age, stage, tumor size, tumor laterality, type of the carcinomatous component (CC), lymph node status, vascular invasion, and degree of surgical debulking. The mean age was 69.0 years (range, 43 to 89 y). The tumor was located in the left and right ovary in 18 and 24 patients, respectively (laterality could not be determined in 5 cases). The mean tumor size was 13.6 cm. Surgical debulking was optimal in 28, suboptimal in 6, and unclear in 13 patients. FIGO stage was I in 1 patient, II in 5 patients, III in 40 (IIIA in 1, IIIB in 11, IIIC in 28), and IV in 1 patient. Node metastasis and vascular invasion were noted in 6/17 and 29/47 patients, respectively. The mean percentage of SC was 29% (median 20%; range, 1% to 90%). The SC was heterologous in 34 (72%) and homologous in 13 (28%) patients. The mitotic figures per 10 HPF in SC were 33 (0 to 128). Tumor necrosis was present in 45/47 cases (mean 10%; range, 1% to 40%, only in CC in 14, only in SC in 2, in both SC and CC in 29). The CC was high-grade serous in 27 patients, endometrioid in 2, mixed high-grade serous and endometrioid in 17, and mixed high-grade serous and clear cell carcinoma in 1 patient. The extraovarian tumor contained only CC in 17 cases, only SC in 1 case, and both SC and CC in 28 cases. The median follow-up was 29 months (range, 1 to 183 mo): 6 patients were lost to follow-up, 3 died postoperatively, 29 died from disease, 2 died from other causes, and 7 were still alive (14 to 183 mo). The DSS rate at 1, 2, and 5 years was 75%, 56%, and 21%, respectively. Presence of SC outside the ovary was a significant adverse prognostic factor (P=0.03), whereas other parameters were not. After adjusting for FIGO stage, presence of SC outside the ovary was still a significant adverse prognosticator for stage III patients (P=0.003), whereas others were not. Therefore, our findings showed that presence of SC outside the ovary was a significant adverse prognostic factor. We advocate listing the specific extraovarian tumor component (SC and/or CC) in the pathology report for primary ovarian malignant mixed mesodermal tumors.