The purpose of this research is to develop formulations to measure the testability of a program. Testability is a program's property that is introduced with the intention of predicting efforts required for testing the program. A program with high degree of testability indicates that a selected testing criterion could be achieved with less efforts and the existing faults can be revealed more easily during testing. In this paper we propose a new program normalization strategy that makes the measurement of testability more precise and reasonable. If the program testability metric derived from data flow analysis could be applied at the beginning of software testing phase, much more effective testing resources allocation and prioritizing is possible.