Verification of complex behavioral models has become a critical and time-consuming process. Determine when to switch to different testing strategy phase is key to improving efficiency. This paper presents an overview of the existing statistical stopping rules that can be used for behavioral models verification. We examined the stopping rules using two VHDL models for five consecutive test phases. The results of the coverage gained and the number of testing patterns applied are then compared for each stopping rule. We conclude that the confidence-based stopping criterion out-performs others in terms of efficiency.