Primary central nervous system lymphoma (PCNSL) is a rare and aggressive brain tumor with an unsatisfactory prognosis. This is probably caused by the little knowledge of the biology and the molecular mechanisms of the disease with a consequent lacking of new targeted therapies and by the poor clinical conditions and performance status of patients, rendering their enrollment in prospective trials very difficult. Chemotherapy followed by radiotherapy is the most commonly used strategy for patients with PCNSL, which is not only associated with better efficacy rates, but also with high incidence of severe neurotoxicity. A relevant dilemma in PCNSL treatment regards the choice between strategies designed to intensify therapy to improve the cure rate, versus strategies of treatment deescalation to avoid severe neurotoxicity. The efficacy of chemotherapy is strongly limited by the special functional and microenvironmental characteristics of the central nervous system (CNS), which is variably protected by the blood–brain barrier (BBB) and includes extensive chemotherapy sanctuaries where tumor cells grow undisturbed. Methotrexate plus cytarabine combination is the current standard chemotherapeutic approach for newly diagnosed PCNSL, since it is supported by a recently published randomized trial. Other strategies are based on the use of other chemotherapy agents (temozolomide, topotecan, and thiotepa) and monoclonal antibodies (rituximab). Consolidation after chemotherapy represents the best role for radiotherapy. Some authorities are investigating in randomized trials which impact on outcome and neurotolerability of replacing consolidation radiotherapy with other strategies such as high-dose chemotherapy (HDC) supported by autologous stem cell transplantation (ASCT).