Glioblastoma, a primary brain tumor, is resistant to chemotherapy and can develop into a fatal space-occupying lesion. Glioma stem cells (GSCs) are thought to be responsible for tumor growth, chemo-resistance, and recurrence. Clonal glioma sphere (GS) culture, in which GSCs are enriched and self-renew as GS clone populations, provides us with quantitative details regarding GS clone survivability and changes in growth during GS/GSC population self-renewal. Previously, we proposed a novel chemotherapeutic paradigm, temozolomide (TMZ) and OTS964 in combination (T&O), and showed that T&O efficiently eliminated self-renewing GS clones and significantly suppressed the regrowth of TMZ-sensitive GS clones. However, it remained unclear whether T&O would be effective in treatment of TMZ-resistant GSC populations. T&O did not suppress T98-GS clone growth during population self-renewal, suggesting that TMZ-like growth suppression is necessary for the long-term control of GSC population size. In this study, we tested the PI3K inhibitor LY294002, which is thought to suppress GSC self-renewal, alone and in combination with OTS964 (L&O) against T98G-GS populations. LY294002 alone suppressed T98G-GS clone growth for 2–3 weeks, while allowing the clones to survive. By contrast, L&O efficiently eliminated two-thirds of the T98G-GS clones and continuously suppressed T98G-GS clone regrowth for 2–3 times longer than LY294002 alone, suggesting that L&O represents an alternative to T&O. Our findings indicate that this quantitatively validated combination paradigm could control growth of TMZ-resistant GSC populations through immediate and sustained shrinkage of heterogeneous GSC populations.