Cerebellar astrocytomas, as a group, carry a more favorable prognosis than most other brain tumors, because these neoplasms generally are histologically benign and amenable to extensive resection. However, it is clear that a number of factors have an impact on prognosis. In particular, resection extent has been strongly associated with progression-free survival: patients undergoing gross total resection appear to have a substantially better prognosis than those undergoing incomplete resection. Brainstem invasion, which is the factor that most often precludes a complete resection, has also been associated with a less favorable prognosis. In addition, histological features indicative of malignancy are clearly associated with a poor outcome. In contrast to the above observations, which have been established convincingly in the literature, a number of issues regarding cerebellar astrocytomas remain unresolved. First, the correlation between histology and prognosis among patients with low-grade cerebellar astrocytomas is uncertain: in some series, pilocytic astrocytomas have been associated with a better prognosis than non-pilocytic tumors, but in other studies, no such relationship has been observed. Second, the role of radiotherapy after incomplete resection of a low-grade cerebellar astrocytoma remains problematic. In view of the lack of convincing data in this regard, many groups, including our own, defer radiotherapy until there is evidence of progressive disease that is surgically unresectable. Finally, the frequency of follow-up in patients with cerebellar astrocytomas remains largely empirical. Although most recurrences are detected within a few years after initial surgery, late recurrences are well known, which raises the question of when and if such patients should be regarded as "cured' of their disease. Long-term multi-institutional natural history studies are in progress to address the above issues.