Objective : The goal of treatment for spinal cord ependymoma is complete removal without postoperative neurological deficit. The authors analyzed the surgical results and factors influencing the postoperative prognosis. Methods : Fifty-one cases of primary spinal cord ependymoma, surgically treated between 1979 and 2003, were retrospectively analyzed. The mean follow-up period was 44 months. Results : Gross total removal was achieved in 42 patients and incomplete removal in nine. The proportion of complete surgical removals was influenced by tumor location and histology. Disease progression was observed in five cases [9.8%], the mean progression free interval after surgical removal was 48 months and the 5-year progression free rate was 68%. Disease progression was found in none of the 42 cases who underwent complete removal, and in 5 of 9 cases who hadincomplete removal group [P<0.001]. Statistically significant disease-progression factors by multivariate analysis were the surgical extent of removal [P=0.012]. preoperative functional status [P=0.032] the presence of intratumoral cysts [P=0.007] and postoperative radiation therapy [P=0042]. Of those patients who underwent incomplete removal, radiation therapy was found to significantly improve the clinical result [P=0042]. Conclusion : In the surgical treatment of spinal cord ependymoma, preoperative functional status, the presence of intratumoral cysts, the extent of removal, and postoperative radiation therapy were found to be significant prognostic factors of postoperative outcome.