Glioblastoma (GBM) is the most aggressive malignant glioma, with a very poor prognosis; as such, efforts to explore new treatments and GBM’s etiology are a priority. We previously described human GBM cells (R2J-GS) as exhibiting the properties of cancer stem cells (growing in serum-free medium and proliferating into nude mice when orthotopically grafted). Sodium selenite (SS)—an in vitro attractive agent for cancer therapy against GBM—was evaluated in R2J-GS cells. To go further, we launched a preclinical study: SS was given orally, in an escalation-dose study (2.25 to 10.125 mg/kg/day, 5 days on, 2 days off, and 5 days on), to evaluate (1) the absorption of selenium in plasma and organs (brain, kidney, liver, and lung) and (2) the SS toxicity. A 6.75 mg/kg SS dose was chosen to perform a tumor regression assay, followed by MRI, in R2J-GS cells orthotopically implanted in nude mice, as this dose was nontoxic and increased brain selenium concentration. A group receiving TMZ (5 mg/kg) was led in parallel. Although not reaching statistical significance, the group of mice treated with SS showed a slower tumor growth vs. the control group (p = 0.08). No difference was observed between the TMZ and control groups. We provide new insights of the mechanisms of SS and its possible use in chemotherapy.