“Singing to another tune” is from Las Leys d’amors (The Laws of Love), a poetic treatise compiled by Guilhem Molinier in the first half of the fourteenth century. Guilhem’s phrase pertains to a compositional technique known to modern scholars as contrafacture, in which the troubadour fashions new lyrics after the poetic structure of a preexistent song, thereby allowing his work to be sung to the earlier melody. The technique of contrafacture is documented not only by Guilhem and contemporaneous theorists but also by the troubadours themselves, who on a number of occasions acknowledge composing a poem “el so de,” or “to the tune of” another composer. Both theory and practice demonstrate that structural imitation came to be most closely associated with several specific genres, including the sirventes (moralizing piece), tenso (debate song), coblas (song of few strophes), and planh (lament), their poetic structures commonly modeled after those of the canso, the dominant genre of troubadour composition. Despite abundant structural indications of contrafacture within the troubadour repertoire, melodic traces of the practice are surprisingly scant. Confirmation of melodic borrowing depends upon the preservation of a model and its contrafactum with their concordant musical readings, yet the small proportion of surviving troubadour melodies (with only one in ten lyric texts transmitted with its tune) poses a significant impediment to melodic corroboration. Only three sirventes have been preserved with melodies that duplicate those of preexistent cansos. In the remaining instances in which a sirventes, tenso, or other imitative type is preserved with a melodic unicum, scholars of troubadour song have tended to maintain that, absent melodic corroboration, the tune must be presumed original rather than borrowed. In view of the sparseness of the musical record, however, one should give consideration to an alternate interpretation, namely that the tune preserved exclusively with a given troubadour’s sirventes and thereafter transmitted as his invention may actually have been borrowed from a preexistent canso whose melody is no longer extant in its original setting. Isolating viable structural models for such suspected contrafacta allows the possibility of reascribing potentially borrowed melodies to their original composers. The study of contrafacture can thus lead us to question the received attributions of a number of tunes, thereby posing a challenge to the readily made assumption that the manuscript rubrics consistently pertain to both text and melody. By examining several suspected cases of contrafacture within a web of relevant indices– e.g., generic norms, intertextual correlations, socio-historic context, rhetorical motivation, transmission, and melodic style– we gain greater insight into a compositional technique that indelibly marked the art of the troubadours.