There has been considerable controversy over the concept of treating glioblastoma multiforme with cytoreductive surgery. Therefore, a retrospective study of cases treated between 1986 and 1991 was conducted to analyze and compare the results of stereotactic biopsy followed by radiation therapy performed in 58 patients with those of surgical resection plus radiation therapy in 57 patients. In both groups, conventionally fractionated radiation (1.7 to 2.0 Gy/day) was delivered, with a total dose of 50 to 60 Gy. Biopsy was performed only in patients with tumors judged to be inoperable. These patients carried a higher surgical risk and were in worse neurological condition than the patients in the resection group. The median survival time for the resection group was 39.5 weeks, as compared with 32 weeks for the biopsy group. This difference was not significant. The most important prognostic factor was the patient's age. The treatment variable biopsy versus resection did not reach prognostic relevance. In patients with midline shift who underwent biopsy, the Karnofsky Performance Scale score decreased in more patients during radiation therapy. The clinical status 6 weeks after surgery, however, showed no significant differences between the two groups. The comparable survival times for the two groups place doubt on the concept of treating glioblastoma multiforme with cytoreductive surgery. Presently, radiation therapy is the most effective treatment for patients with glioblastoma. There is no question that decompressive surgery followed by radiation therapy should be performed whenever necessary for sever space-occupying lesions and when it will not cause new neurological deficits.