Cancer stem cells (CSCs) form a small proportion of tumor cells that have stem cell properties: self-renewal capacity, the ability to develop into different lineages and proliferative potential. The interest in CSCs emerged from their expected role in initiation, progression and recurrence of many tumors. They are generally resistant to conventional chemotherapy and radiotherapy. There are two hypotheses about their origin: The first assumes that CSCs may arise from normal stem cells, and the second supposes that differentiated cells acquire the properties of CSCs. Both hypotheses are not mutually exclusive, as it is possible that CSCs have a diverse origin in different tumors. CD133+ cells (CD133 is marker of CSC in some tumors) isolated from NBL, osteosarcoma and Ewing sarcoma cell lines are resistant to cisplatin, carboplatin, etoposide and doxorubicin than the CD133-ones. Being resistant to chemotherapy, there were many attempts to target CSCs epigenetically including the use of histone deacetylase inhibitors. The diverse influence of valproic acid (histone deacetylase inhibitor) on normal and cancer stem cells was proved in different experiments. We have found an increase percentage of CD133+ NBL cells after their incubation with VPA in a dose that does not induce apoptosis. Further researches on CSCs and clinical application for their detection are necessary: (i) to define the CSC function in carcinogenesis, cancer development and their role in metastasis; (ii) to find a specific marker for CSCs in different tumors; (iii) to explain the role of different pathways that determine their behavior and (iv) to explain mechanisms of chemoresistance of CSCs.