We studied the influence of the type of surgery (microsurgery or macrosurgery) and extent (complete resection with lobectomy, complete resection alone, partial resection with lobectomy or partial resection alone) on early postoperative results and survival time in 118 consecutive patients who underwent surgery for glioblastoma multiforme. Early results were assessed by the Karnofsky score at 4 weeks postoperatively. Survival was compared using Kaplan-Meier curves and Mantel statistics. The median survival time (MST) after microsurgery (12.1 months) was significantly longer than that after macrosurgery (7.3 months). The longer survival after microsurgery was, however, largely attributable to better early results and a consequently higher proportion of patients who could undergo radiotherapy. Complete resection was superior to partial resection. Additional lobectomy did not appreciably influence the early results and the MST in completely resected tumours. So the MST after complete resection in the microsurgical group without lobectomy was 12.6 months, with lobectomy 12.9 months. In the macrosurgical group the respective values were 7.4 months without and 8.2 months with lobectomy. In incompletely resected tumours lobectomy worsened the early results compared to incomplete resection alone and led to a shorter MST.