Objective: There are many kinds of intracranial lesions in children, and although the application of various medical imaging techniques can improve the diagnosis rate of intracerebral lesions, the phenomenon of “different images of the same disease and different images of the same disease” of intracerebral lesions exists widely, and histopathology is still the gold standard of diagnosis and the basis for pediatric neurologists to decide the subsequent treatment. Stereotactic brain lesion biopsy is the best minimally invasive method to obtain tissue specimens, especially for those who cannot obtain pathological specimens by craniotomy. The purpose of this study was to investigate the significance and experience of stereotactic brain tissue biopsy in the diagnosis of pediatric neurological diseases. Methods: We retrospectively analyzed the medical records of 138 patients under 14 years of age who underwent stereotactic brain tissue biopsy in our department between December 1987 and January 2009, including 46 CT-guided biopsies and 92 MRI-guided biopsies, including 24 hydrogen proton magnetic resonance spectroscopy (1H-MRS)-guided brain biopsies and 18 stereotactic image-guided neuroendoscopy-assisted biopsies. All patients with frame-free stereotactic biopsies were treated with basic intravenous general anesthesia, plastic pillow fixation of the head, and tracheal intubation or laryngeal mask to keep the airway open according to the ventilation status. All directional surgical planning used stereotactic surgical software to reconstruct the lesion in three dimensions and design the target point and puncture retrieval path. RESULTS: Among 138 pediatric patients with intracranial lesions, 84 (60.9%) were male and 54 (39.1%) were female; ages ranged from 1 to 14 years, with a mean of 8.9 years; 88 (63.7%) were single intracranial lesions, 32 (23.2%) were multiple lesions, and 18 (13%) were diffuse hemispheric multileaf infiltrating lesions. The biopsy targets were selected according to the most obvious sites of the lesions on the imaging, including 119 cases of single-target sampling, 14 cases of double-target sampling, and 5 cases of triple-target sampling; 162 targets were distributed as follows: 32 in the frontoparietal-occipital lobe semi-oval center, 14 in the temporal lobe, 37 in the basal ganglia and thalamus, 14 in the saddle area and paraventricular area, 7 in the corpus callosum, 18 in the lateral wall and ventricles of the ventricles, 6 in the cerebellar hemispheres, 25 in the pontine brain, pontine arm, and brainstem, and 9 in the pineal gland area. Various oncological pathological diagnoses were obtained in 104 cases (75.4%), including 65 cases of tumors of neuroepithelial origin, 28 cases of germ cell tumors, 4 cases of leukemic intracerebral invasion, 4 cases of primary lymphoma, 1 case of pineal cell tumor, 1 case of melanoma, and 1 case of yolk sac tumor. Among the 65 tumors of neuroepithelial origin: astrocytoma and oligodendroglioma 46 cases; 8 cases of supratentorial primitive neuroectodermal tumors; 3 cases of embryonic dysplastic neuroepithelioma; 4 cases of ganglioglioma or ganglioglioma; 1 case of medulloblastoma; and 4 cases of ventricular meningioma. The remaining 26 non-neoplastic lesions included 2 cases of viral encephalitis, 11 cases of multiple sclerosis and verrucous demyelinating disease; 4 cases of neurodegenerative diseases (including 2 cases of mitochondrial encephalomyopathy and 2 cases of metabolic encephalopathy); 1 case of central nervous system vasculitis; 1 case of tuberculoma, 4 cases of parasitic granuloma, and 3 cases of infectious granuloma and small abscess. The histopathological diagnosis could not be clarified in 8 cases (5.79%), and the total positive biopsy diagnosis rate reached 94.20%. One case of directly related death due to biopsy hemorrhage was a biopsy of a lesion in the pineal region (0.7%), and 6 cases of small hematoma formation without neurological deficits due to a small amount of blood leakage from the target site (hematoma volume