Prosodic properties are features of human speech sound that can be assigned to units larger than a segment like syllables and phrases. Earlier studies have established that not all components of the speech signal are perceived to be distinctive by speaker-hearers. It is not enough therefore to establish the acoustic correlates of “stress” if the same cannot be so identified through simple hearing. This study aims to establish by experiment and statistical methods the relative contribution of “stress” variables (pitch, length, and intensity) in distinguishing one member of minimal pairs or set from the other. Through GoldWave and PRAAT programs, the variables are modified and sets of synthesized utterances are evaluated by a sample of native speakers. By comparing the results of this experiment with those acoustic measurements the researchers intend to determine with confidence the variable or combination of variables that dictates what has been widely referred to as stress.