The musical analysis of Greek tragedy has traditionally been limited to studies of meter and metatheatrical language. This dissertation seeks to establish a new approach to ancient dramatic song by demonstrating that the linguistic pitch accents of tragic lyrics often trace the melodic contours of their lost musical settings. In the papyri and inscriptions that preserve music notation alongside Greek lyrics, intonation and melody are often coordinated according to set principles, which are well established by previous scholarship. Through the creation of software that applies these historical principles to tragic texts, I demonstrate that stanzas sung to the same melody are significantly more similar in their accentual contours than control groups that do not share a melody. In many instances, the accents of these paired texts consistently trace the same pitch contours, allowing us to reconstruct the shape of the original melody with a high degree of confidence.After a general introduction, the dissertation’s first two chapters address the historical basis for this approach. Chapter 1 reviews the evidence for the musical structure of tragic song, confirming the widely held view that paired stanzas were generally set to the same melody. Chapter 2 turns to the evidence for the role of pitch accents in ancient Greek song, including the ancient testimony and musical documents, and a computational study of accent patterns across all the lyrics of Aeschylus’ surviving tragedies. The methodology developed in these first two chapters is applied in two case studies, in which I reconstruct and interpret the accentual melodies of select tragic lyrics. Chapter 3 analyzes the musical design of the chorus’ entrance song in Aeschylus’ Agamemnon, along with sections of the Kommos from Choephori. In both cases, I argue, melody would play an integral role in highlighting the themes of repetition and reversal within the Oresteia. Chapter 4 turns to the music of Euripides’ Medea, a play that has been central to previous discussions of accent in tragic music. Reading the lyrics and accentual melodies within the framework of musical history as understood in the fifth century bce, I argue that Euripides uses a contrast between ‘old’ and ‘new’ melodic styles to position his chorus at a turning point within literary history. In the dissertation’s final chapter, I address the reception of Medea’s music in a fragmentary comedy, the so-called Alphabet Tragedy of Callias. Together, these interpretive chapters provide a template for future work applying methods of musical analysis to the accentual melodies of ancient Greek song.