While musical sources and documents from throughout the Middle Ages reveal that mode was an enduring and consciously derived trait of monophonic chant, modality in later polyphony shares neither the historical span nor the theoretical clarity of its monophonic counterpart. Modern theorists are left with little more than circumstantial evidence of the early development of modality in polyphony. This study attempts to shed light on the problem by detailed analysis of a select body of paraphrase masses from the early sixteenth century. First, it correlates the correspondence between the paraphrased voice and the original chant, establishing points of observation that become the basis of melodic analysis. Then, these points are correlated with known rules of counterpoint. Exceptions are identified and examined for their potential to place emphasis on individual mode-defining pitches. A set of tools is derived for quantifying the relative strength of cadential actions. Levels of cadence are defined, ranging from full, structural cadences to surfacelevel accentuations of individual pitches by sixth-to-octave dyadic motions. These cadence levels are traced through the Missae de beata virqine repertoire from c. 1500-1520, a repertoire that includes masses of Josquin, Brumel, La Rue, Isaac, and Rener. While the Credos, based on two chant sources—one early (11th century) and one later (15th century)—showed little modal consistency, the Kyries show some suggestion of purposeful modal expression; and the Glorias show even greater implications. Results of the study have potential application in sixteenth-century music scholarship to such important issues as musica ficta, performance practice, text underlay, and form.