Approximately 20% of children with acute lymphoblastic leukemia (ALL) relapse because of failure to eradicate the disease. Current drug efficacy studies focus on reducing leukemia cell burden. However, if drugs have limited effects on leukemia-initiating cells (LICs), then these cells may expand and eventually cause relapse. Parthenolide (PTL) has been shown to cause apoptosis of LIC in acute myeloid leukemia. In the present study, we assessed the effects of PTL on LIC populations in childhood ALL. Apoptosis assays demonstrated that PTL was effective against bulk B- and T-ALL cells, whereas the CD34(+)/CD19(-), CD34(+)/CD7(-), and CD34(-) subpopulations were more resistant. However, functional analyses revealed that PTL treatment prevented engraftment of multiple LIC populations in NOD/LtSz-scid IL-2Rγ(c)-null mice. PTL treatment of mice with established leukemias from low- and high-risk patients resulted in survival and restoration of normal murine hemopoiesis. In only 3 cases, disease progression was significantly slowed in mice engrafted with CD34(+)/CD19(-) or CD34(+)/CD7(-) and CD34(-) cells, but was not prevented, demonstrating that individual LIC populations within patients have different responses to therapy. These observations indicate that PTL may have therapeutic potential in childhood ALL and provide a basis for developing effective therapies that eradicate all LIC populations to prevent disease progression and reduce relapse.