Albion and Albanius (1685) and King Arthur (1691) were simultaneously conceived by the poet and playwright John Dryden in view of the 25th anniversary of the Restoration. While the former work was originally intended as an allegorical prologue to King Arthur, it was inflated into an independent, all-sung opera and set to music by Louis Grabu. Its failure in 1685 caused its tandem piece to be shelved at the time. Six years later, King Arthur was produced after all, as a dramatick opera combining spoken dialogue with musical episodes by Henry Purcell. In his preface to Albion and Albanius, Dryden had developed a tentative theory of libretto writing, by distinguishing between different vocal styles and their respective diction, metre and rhyme. After introducing this theoretical model, the present paper analyses King Arthur along similar lines. Do the musical lyrics in this dramatick opera follow the structural principles outlined in Dryden’s libretto model? How does Purcell’s setting confirm or conflict with the musico-dramatic blueprint designed by his librettist? And how should we assess Dryden’s prefatory complaint that “the Numbers of Poetry and Vocal Musick, are sometimes so contrary, that in many places I have been oblig’d to cramp my Verses, and make them rugged to the Reader, that they may be harmonious to the Hearer”? In the absence of Dryden’s autograph, it is only through the structural analysis of the printed word-book and its musical setting that we can begin to retrieve the meaning of concepts like ‘cramped’ or ‘rugged’ verses – often quoted but still arcane. A musico-poetic analysis of King Arthur against the backdrop of Dryden’s theoretical groundwork sheds new light on the literary material of Restoration opera and on the creative dialogue between lyricists and composers at the time.