The results of adjuvant chemotherapy in high-grade gliomas are disappointing. Carmustine (BCNU) wafers placed in the resection cavity provide a modest increase in median survival. However, the overall benefit of this approach is small, it offers no potential for cure, and the results are not statistically significant in patients with glioblastoma multiforme. During the past three decades, controlled clinical trials of systemically administered adjuvant chemotherapy have consistently failed to show significant benefit in patients with glioblastoma multiforme. Furthermore, as of 2003, no randomized studies convincingly document improved survival from adjuvant chemotherapy in anaplastic astrocytomas or anaplastic oligodendrogliomas, which are the most chemotherapy sensitive gliomas. These findings result from the limited efficacy of available chemotherapy in malignant gliomas and the current design of adjuvant brain tumor trials that allow accrual of patients with bulky residual disease. Until further information becomes available, it is reasonable to use radiation alone as adjuvant therapy and to reserve chemotherapy for the treatment of recurrences. This approach minimizes toxicities and ensures that chemotherapy is administered when its effect can be assessed using measurable disease. Improving outcome in high-grade gliomas requires an emphasis on discovering novel therapies and improving the design of adjuvant trials.