The increasing penetration of renewable energy (RE) brings nonnegligible uncertainties to a power system, which should be carefully considered in the day-ahead scheduling for better RE utilization and higher power supply reliability. However, stochastic unit commitment of multitype power is a challenging problem considering the uncertainties of RE and the complicated characteristics of different power. This paper specifically focuses on the modeling and optimization approach for multitype power sources stochastic unit commitment (MPSSUC) with high penetration of RE and multiple uncertainties using interval number programming (INP). The MPSSUC model is established considering the elaborate characteristics of thermal power, hydropower, wind power and pumped storage power. The uncertainties of wind energy, natural water inflow, and power load are depicted by interval numbers. A novel particle swarm optimization–based bilevel solving approach is proposed for MPSSUC optimization, which preserves the interval properties of INP for better accommodation of uncertainties. Case studies on the IEEE 118-bus system and a realistic power system show that this study can effectively improve the RE accommodation while maintaining operation benefit. Analyses on the uncertainty level influence, trade-off between cost and robustness, and the operation characteristics by water regimens are also presented.