Monte Carlo simulations have been widely adopted for analyzing the circuit properties, such as SRAM yield, under the influence of process variations. Enormous calculation time is required in such a simulation because of low probability of defective SRAM cells. In this paper, we propose a robust shift-vector determination for mean-shift importance sampling, by which efficiency and stability of the Monte Carlo simulation is significantly improved. In the proposed method, the hypersphere sampling is developed to find the optimal shift-vector autonomously. The sampling is also limited to the regions where meaningful contribution to the yield is recognized. Simulation examples reveal that the proposed technique stably and efficiently estimates yield of noise stabilities of an SRAM cell. At the failure probability of 10^(-10), the number of calculation trials has been reduced by six-orders magnitude compared with the conventional Monte Carlo simulation.