With computed tomography (CT) and magnetic resonance imaging stereotactic systems, biopsies of intracranial lesions can be made with safety and ease. Before the development of this technique, neurosurgeons often performed freehand brain biopsies under CT guidance. While stereotactic biopsy is the procedure of choice for small, deep lesions, few studies have compared the morbidity, mortality, and efficacy in obtaining a diagnosis associated with these two techniques for superficial lesions. A total of 167 consecutive CT-guided or stereotactic brain biopsies were performed in 154 patients. Fourteen of the stereotactic and 12 of the CT-guided biopsies were of deep lesions and were excluded from analysis. The results of 75 freehand CT-guided biopsies of superficial lesions in 69 patients were compared with those of 66 stereotactic biopsies (34 CT-guided and 32 MRI-guided) performed with the Brown-Roberts-Wells stereotactic system in 60 patients. Twenty-five of the lesions in the stereotactic biopsy group measured < or = 2 cm, as compared with 13 of those in the freehand CT-guided biopsy group. There were no biopsy-related deaths among the patients who underwent freehand CT-guided biopsy and one death among those who underwent stereotactic biopsy (1.5%). Freehand CT-guided biopsy was associated with 5% morbidity, compared with 6% morbidity for stereotactic biopsy. Seven CT-guided needle biopsies (9%) and 12 stereotactic biopsies (18%) were nondiagnostic. Statistical analysis showed no significant difference between morbidity and mortality in the two groups, but the rate of diagnostic failure was significant (P < 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)