The purpose of this research was to determine whether there were significant differences in male or female middle-school students' ability to appropriately convey the emotions of happiness, sadness, anger, and fear while singing musical passages within the context of four treatment situations. Four treatment methodologies were designed to identify whether (a) participants can be taught to sing with emotional expressiveness, and (b) a specific treatment method would be significantly better for this task. The treatments used four different learning conditions: traditional instruction, aural modeling, kinesthetic exploration, and audio-visual learning. Each treatment enabled participants to explore different approaches to performing music with emotion. A significant interaction effect was found between males and females in their rated ability to convey anger in their voices, with females scoring significantly lower than males. Happiness and sadness were conveyed equally well by males and females, and fear was difficult for both to convey. There was a statistically significant difference found in effectiveness of the four treatment methodologies, showing modeling (treatment II) and audio-visual (treatment IV) methodologies to be significantly better than verbal instruction (treatment I). Implications for music education and directions for future research were discussed. (Contains 2 tables and 1 figure.)