PURPOSE: In spite of advances in therapy for some subtypes, group 3 medulloblastoma continues to portend a poor prognosis. A subpopulation of medulloblastoma cells expressing the cell surface marker CD133 have been posited as possible stem cell like cancer cells (SCLCC), a potential source of drug resistance and relapse. Retinoids have been shown to affect SCLCC in other brain tumors. Based on these findings, we hypothesized that the CD133-enriched cell population group 3 medulloblastoma cells would be sensitive to the novel rexinoid, UAB30. METHODS: Human medulloblastoma cell lines were studied. Cell sorting based on CD133 expression was performed. Both in vitro and in vivo extreme limiting dilution assays were completed to establish CD133 as a SCLCC marker in these cell lines. Cells were treated with either retinoic acid (RA) or UAB30 and sphere forming capacity and CD133 expression were assessed. Immunoblotting was used to assess changes in stem cell markers. Finally, mice injected with CD133-enriched or CD133-depleted cells were treated with UAB30. RESULTS: CD133-enriched cells more readily formed tumorspheres in vitro at lower cell concentrations and formed tumors in vivo at low cell numbers. Treatment with RA or UAB30 decreased CD133 expression, decreased tumorsphere formation, and decreased expression of cancer stem cell markers. In vivo studies demonstrated that tumors from both CD133-enriched and CD133-depleted cells were sensitive to treatment with UAB30. CONCLUSIONS: CD133 is a marker for medulloblastoma SCLCCs. Both CD133-enriched and CD133-depleted medulloblastoma cell populations demonstrated sensitivity to UAB30, indicating its potential as a therapeutic option for group 3 medulloblastoma.