Statistical structure modeling can help to establish the glass formulation design models to meet the vitrification requirements of different types of high-level waste liquids due to its advantages of high efficiency and accuracy. Taking the glass transition temperature Tg, thermal expansion coefficient α and element leaching rate of Li, Na, B as target properties, we utilized a statistical structure simulation method for the development of nuclear waste glass curing formula. The results show that Tg, α and chemical stability of the glasses can be well simulated with the statistical structural data. It is also indicated that the data predicted by the model are in a reasonable agreement with the measured results at a simulation desirability 0.94. Statistical structure modeling can assist to establish the formula database of the high-level liquid waste vitrification.