In late imperial China, historical narratives flourished across multiple genres of text, as authors were increasingly breaking free of the boundaries of standard historiographical forms and placing information in fictional and dramatic texts. The advent of large open-access digital corpora offers increasingly representative samples of imperial literary production, which I use to systematically study this phenomenon. In this paper, I algorithmically extract descriptions of historical events from a corpus consisting of historical, fictional, and dramatic texts dating from mid-Ming to high-Qing China (approximately 1500 to 1800). I track when imperial texts mention dates and people, extract when works are textuality intertwined, and measure information density. I couple this with large-scale bibliographic analysis based on public domain library catalog records and contemporary bibliographies to map the historical focus of late imperial writing. I show that historical attention varies in genre specific ways and is dependent on myriad literary and social factors.