The paper is an attempted contribution to the study of the Thousand and One Nights in terms of genre and formula. The object of investigation is the story of “King ‘Umar ibn al-Nu’m?n and His Family.” The longest tale in the Nights, it takes up approximately one-eighth of the entire work, and contains material of diverse generic provenance. Although manuscript evidence suggests that it was included in recensions of the Nights from a relatively early date, it is nonetheless distinct from much of the collection in its resemblance to the much lengthier popular epics. Its interest and appeal may in fact lie in its deviation from genre standards. In some respects it is s?ra (e.g., its relative length, formulaic conventions, certain character types; interlacing of narrative strands as opposed to embedding); in others it is typical of the Nights’ hik?y?t (less rhymed prose, romance themes, embedded stories, pious and learned digressions). Some of these formal differences between hik?ya and s?ra are well enough known to be taken for granted. In my paper I examine how these differences function in terms of plot and narrative discourse (especially voice and mood). Drawing on the works of those who have treated the epic in a comparative context (Bakhtin, Goody, Fusillo), I demonstrate how the two genres of hik?ya and s?ra coexist in the story of ‘Umar al-Nu’m?n, and specifically how each rests on a fundamentally different conception of knowledge. The intricately-plotted core tales of the Nights are clearly products of a learned milieu, their knowledge corresponding to that of the scholars and littérateurs, i.e. a subjective mental process or a fixed body of factual material. In general the s?ra is less clever in terms of plotting, but is suffused with a much more volatile concept of what can be known with certainty. Both of these narrative epistemologies are at work in the tale of ‘Umar al-Nu’m?n.