We have recently described a technique allowing routine integration of PET scan data in the planning of stereotactic brain biopsy [Levivier et al:, Neurosurgery, 1992; 31:792-797]. We now report our results in a consecutive series of 38 patients that underwent combined FDG-PET and CT-guided stereotactic biopsy between June 1991 and January 1993. Stereotactic procedures were performed according to a standard protocol in which two biopsy trajectories were used whenever possible. This yielded to a total of 78 trajectories. In 31 patients, at least 1 trajectory was defined using an abnormality visualized on PET. The histological diagnosis was obtained in all cases. The diagnostic yield of each trajectory according to the FDG uptake of PET and contrast enhancement on CT showed that most low-grade tumors were found in hypometabolic/hypodense areas, glioblastomas were all diagnosed in areas with increased FDG uptake and contrast enhancement on CT while data for anaplastic astrocytomas were heterogeneously distributed. In the present series, 6 targets defined on CT were nondiagnostic; this never occurred when targets were defined on FDG-PET. In one case of anaplastic astrocytoma, the diagnosis was only made using a target defined on PET in a normal CT area; the second trajectory was made in a contrast-enhanced area on CT and was nondiagnostic. Because only structural images such as CT or MRI are available preoperatively, we analyzed separately the benefit of FDG-PET-guided biopsy in patients who have either an enhanced- or a nonenhanced lesion on CT. CT-enhanced lesions were found in 27 patients.(ABSTRACT TRUNCATED AT 250 WORDS)