Evolutionary structural testing is a technique that uses specific approaches based on guided searches algorithms. It involves evaluating fitness functions to determine whether test data satisfy or not various structural testing criteria. For testing multi-way decision constructs the nested \textit{If-Then-Else} structure and Alternative Critical Branches (ACBs) approaches are generally used. In this paper a new evolutionary structural approach based on Compact and Minimized Control Flow Graph (CMCFG) which uses two different formulas for evaluating the performance of test data, is presented. The CMCFG approach is derived from the concept of Control Flow Graph (CFG). Experiments on different \textit{Switch-Case} constructs with different nesting levels have demonstrated that CMCFG yields significantly better results in finding test data which cover a particular target branch in comparison with the previous approaches.
