The rational management of intracranial lesions should be based on the exact definition of the nature of the lesions and, when it is possible, on their spatial definition. Since External Radiotherapy (ERT) and cytostatic therapy are not free of undue effects, especially in children, such treatments should be used only when a "sure" diagnosis is obtained. The aim of this paper is to study the results allowed by the Talairach's stereotactic methodology in children. During the period January 1979-December 1986, 704 stereotactic procedures including serial biopsies, were performed at the S. Anne Hospital in Paris. One hundred forty-eight procedures (21%) concerned 134 children (78 M; 56 F) aged from 2 to 16 years. The interval between the occurrence of the first clinical symptoms and the stereotactic procedures varied between 1 and 180 months (m: 24 m). Fifty-two (40%) had previous therapeutic procedures without precise diagnosis. The lesions were hemispheric in 46 (34%) and deep seated in 88 (66%). The serial stereotactic biopsies proved the existence of a non-tumoural lesion in 20 children (14.9%): (cryptic vascular malformation: 5, cortical dysplasia: 3, haematoma: 3, ischaemia: 1, granuloma: 1, degenerative pathology: 2, cicatrix: 2, post-ERT alterations: 1, arachnoidal cyst: 2). Four were in the brain stem. In 3 patients (2%), a precise diagnosis was not obtained. The CT scan characteristics of the 20 non-tumoural lesions did not permit to establish a precise differential diagnosis. The therapeutic management was adapted to the diagnosis, avoiding potentially dangerous procedures in the 20 non-tumoural lesions.