The correlation between the histological features and clinical outcome remains poor in pediatric intracranial ependymomas. We performed a retrospective study of a group of 31 patients (diagnosed from 1985 to 1995) to assess prognostic implications of the current grading system, of histological and immunohistochemical features, and of ploidy status estimated by flow cytometry. Immunoexpression of a broad spectrum of antigens was evaluated, including MIB-1, topoisomerase-IIalpha, cyclin D1, glial and epithelial proteins (GFAP, EMA, cytokeratins), molecules involved in controlling apoptosis (bcl-2, caspase-3/CPP32), and p53 oncoprotein. Univariate and multivariate statistical analyses were performed to evaluate the influence of each variable on both the progression free survival (PFS) and the overall survival (OS) with at least 7-year follow up. Although we showed a significant correlation between histological grade and prognosis, the current grading system failed in predicting outcome in nearly one third of individual cases. Problems with interpathologist reproducibility were also demonstrated. The extent of surgical resection was the only clinical factor that was associated with survival. Both the PFS and the OS were significantly decreased for the following pathological variables: increased cellularity (>300 nuclei per HPF), mitotic activity of >7 per 10 HPF, increased MIB-1 labeling index (LI), topoisomerase-IIalpha LI, S-phase fraction, and p53 and bcl-2 positivity. Increased cyclin D1 LI was demonstrated to have only a marginally significant impact on PFS. A flow chart modeling was further performed to formulate a scheme for discriminating of prognostic subgroups. Based on that, p53 immunopositivity and/or MIB-1 LI of >5% (after subtotal resection) or MIB-1 LI of >15% (after complete resection) were the strongest indicators of the tumor's aggressive behavior and of a poor prognosis of the disease. Foci of hypercellularity should be specifically looked for in ependymomas for assessing the immunohistochemical studies.