Primary CNS germ cell tumors present several biologic and therapeutic concerns. Their predilection to midline locations (pineal and suprasellar) has discouraged until recently attempts at biopsy, much less radical surgical resection. Their heterogeneity (germinomas, embryonal carcinoma, endodermal sinus tumor, choriocarcinoma and malignant teratoma) has made uniform treatment planning unrealistic. Their rarity (less than 50 cases per year in USA) precludes a large institutional experience. More recently, with the use of microsurgical techniques, major resections are possible and accurate pretreatment histologic diagnosis can be established. Following surgery, patients should be staged with cerebrospinal fluid (CSF) cytology, serum and CSF tumor markers (alpha-fetoprotein and human beta-chorionic gonadotrophin) and myelography. The majority of patients will present with disseminated disease and require craniospinal radiotherapy. Chemotherapy should eventually prove to be an effective adjuvant treatment modality, not only for the highly radiosensitive germinoma with a good prognosis (5-year survival greater than 60%) following radiotherapy alone, but also for the radioinsensitive nongerminoma germ cell variants where longterm survival is unusual. Chemotherapy agents such as cyclophosphamide, cisplatin, vinblastine, VP-16 and bleomycin appear to be useful for patients with newly diagnosed and recurrent disease.