This article argues that decades before Sir Walter Scott published his first historical novel, a number of innovative British novels questioned the possibility of rendering historical topics in narrative form. Although Henry Fielding had claimed that the novel should closely resemble historiography, three eighteenth-century British novelists recognized the problems besetting the attempt to provide true stories of the past in different ways. Sarah Fielding tried to get rid of the elevated position of the ‘omniscient’ narrator, rejecting even chronological order in her fictionalised account of the lives of Cleopatra and Octavia. Laurence Sterne parodied not only historiographical ways of arriving at true knowledge of past events in his seminal novel Tristram Shandy, he also demonstrated that the story is the result of the narrator’s use of narrative conventions and of language, and that the gap between the past and the representation of the past cannot be bridged. Maria Edgeworth tried to get rid of the historiographical dilemma by choosing an unreliable narrator as a biographer in her work Castle Rackrent, which is one of those eighteenth-century novels that casts doubt on whether it is possible to tell a true story of the past with those narrative means which Scott would later use to great acclaim.