Testing against historical data is an important step in the validation of simulation models. Because of their wide scope and coverage, global change models require a large amount of data for testing. This update of an earlier version of a hundred-year database used to test global change models reports on data on population, gross domestic product, added value of industry and services, land use and, livestock, as well as on specific data categories for the systems: energy/economy, atmosphere/ocean and terrestrial environment. Where possible, data are organised according to country for the period 1890 - 1990. Some of the data show geographic detail.