Primary CNS lymphoma is an enigmatic and challenging disease—its uncertainties begin with aetiology and extend to the potential for treatment-related toxicity years later. The malignant lymphocytes' tropism for the CNS remains unexplained, both in terms of disease origin and the low incidence of systemic relapse. Presentation ranges from many months of intermittent vague neurological symptoms to rapidly progressive severe sensorimotor and cognitive impairment. Pathological diagnosis can be a problem and hampered by the empirical use of corticosteroids. In The Lancet today, Andrés Ferreri and colleagues from the International Extranodal Lymphoma Study Group (IELSG) present only the second randomised trial reported in primary CNS lymphoma. The investigators conclude that the addition of high-dose cytarabine to high-dose methotrexate improved disease response. In this study, 77% (33/43) of responding patients also received whole-brain irradiation as part of their primary therapy. The addition of cytarabine was also associated with improved overall survival, which confirmed the benefit suggested in the retrospective analysis by the same study group. The investigators are to be commended for their contribution to the evidence-base in this rare disease. Successfully completing clinical trials in common cancers is difficult enough, and opinions on treatment in primary CNS lymphoma are polarised. Patients' median age is about 60 years with many older than 70 years. This age group is historically under-represented in clinical trials and, in patients with primary CNS lymphoma, at highest risk of treatment-related complications.