Montesquieu never referred to Lettres persanes (Persian Letters) as a novel until "Quelques remarques sur les Lettres persanes," which begins: "Nothing about the Lettres persanes was more ingratiating than to find in it unexpectedly a sort of novel. There is a visible beginning, development, and ending […]." Initially, for most of its first readers as well as for its author, it was not considered primarily a novel, and even less an "epistolary novel" (as it is often classified now), which was not at that time a constituted genre. Indeed, it has little in common with the sole model at the time, Guilleragues’s Lettres portugaises of 1669. A collection of "letters" in 1721 would more likely evoke the recent tradition of essentially polemical and political periodicals, such as Lettres historiques (1692–1728), the Jesuits’ famous Lettres édifiantes et curieuses (1703–1776), not to mention Mme Dunoyer’s Lettres historiques et galantes of (1707–1717) which, in the form of a correspondence between two women, provide a chronicle of the end of the reign of Louis XIV and the beginning of the Regency. The Lettres persanes thus helped confirm the vogue of a format that was already established. But it is in its numerous imitations – such as Lettres juives (1738) and Lettres chinoises (1739) of Boyer d’Argens, Lettres d’une Turque à Paris, écrites à sa sœur (1730) by Poullain de Saint-Foix (published several times in conjunction with Lettres persanes), and perhaps especially Françoise de Graffigny’s Lettres d’une Péruvienne (1747) – not to mention the letter-novels of Richardson – which, between 1721 and 1754, had in effect transformed Lettres persanes into an "epistolary novel." Whence this remark in Montesquieu’s Mes Pensées: "My Lettres persanes taught people to write letter-novels" (no. 1621).
The epistolary structure is quite flexible: nineteen correspondents in all, with at least twenty-two different recipients. Usbek and Rica by far dominate with sixty-six letters for the former and forty-seven for the latter (of the final 161). Ibben, who functions more as addressee than correspondent, writes only two letters but receives forty-two. Likewise, an unnamed person (designated only as ***) – if always the same – receives eighteen letters and writes none at all. There is even one complete anomaly, a letter from Hagi Ibbi to Ben Josué (Letter 37 ), neither of whom is mentioned elsewhere in the novel.
The letters are apparently all dated in accordance with a lunar calendar which, as Robert Shackleton showed in 1954, in fact corresponds to our own, by simple substitution of Muslim names, as follows: Zilcadé (January), Zilhagé (February), Maharram (March), Saphar (April), Rebiab I (May), Rebiab II (June), Gemmadi I (July), Gemmadi II (August), Rhegeb (September), Chahban (October), Rhamazan (November), Chalval (December).