This study sought to identify and then quantify vocal parameters that contribute to emotional prosody in order to select appropriate audio stimuli for PROID, a computerized emotion identification task. Phase I: Audio recordings were collected from 6 male and 6 female college students, who read sentences previously established as having neutral content. Sentences were recorded with happy, sad, angry, and fearful emotional prosody each with mild, moderate, and extreme valences. The parameters that were measured for each recording were the pitch, pitch range, and intensity. To compensate for natural vocal differences between subjects, all parameters were calculated as a change from a neutral baseline. We found that each emotion could be uniquely profiled depending on which parameters are affected at each valence, with few differences between the sexes. Phase II: Audio recordings were selected based on how well the changes in vocal parameters matched the previously determined emotional profiles. 80 selected recordings served as stimuli in PROID. The 80 stimuli test was then validated to further determine which recordings were the best examples of emotionally prosodic speech. This task will be used to test for deficits in emotional prosody associated with a wide range of neuropsychiatric disorders. It will also be converted to a functional magnetic resonance imaging (fMRI) task to study areas of activation in the brain when listening to affective speech.