Li,
Epidemic represent high-impact events within the spectrum of human catastrophes, and China possesses an exceptionally rich historical record of such occurrences. Despite the profound impact of epidemics on human history, comprehensive, spatially-explicit long-term datasets are scarce. This study presents the Chinese Historical Epidemic Dataset (CHED), detailing its construction methodology, data architecture, and core features. CHED compiles epidemic records from 1056 BCE to 1949 CE, constituting the most comprehensive dataset on historical epidemics in China to date. Each entry has undergone multiple rounds of sampling, cross-verification, and rigorous textual validation, capturing six key dimensions: time, region, cause, effect, category, and response. The dataset provides annual temporal resolution and county-level spatial granularity. Structured as a hybrid of semi-structured narrative entries, structured time-series records, and digitized spatiotemporal coordinates, CHED supports source traceability, multi-granular querying, and precise spatiotemporal localization. This dataset offers robust empirical support for research into the historical globalization of infectious diseases and the long-term evolution of the environment–health interactions.