Some stats describing the National Archives of Australia collection
What’s actually in the National Archives of Australia? And how much of that has been described, digitised, or opened to the public? I’ve made a few attempts at answering these questions over the years by harvesting summary statistics about every series in RecordSearch. I’ve now saved all this data in Zenodo. The dataset description reads:
This dataset contains information and statistics describing record series held by the National Archives of Australia (NAA). The data was harvested from RecordSearch, the NAA’s online database, using this notebook.
Under the Commonwealth Record Series System a ‘series’ is defined as ‘a group of records that has resulted from the same accumulation or filing process or that has a similar format or information content’. Series are created by government agencies, and contain any number of items. The NAA uses series to organise, describe, and manage its holdings. For example, Series A1 contains over 60,000 correspondence files created by the Department of External Affairs and its successors between 1903 and 1938.
This dataset aims to provide an overview of the NAA’s holdings by compiling basic information about each series. It contains four data harvests created in late 2016, May 2021, April 2022, and May 2025. By comparing these different data files it is possible to observe changes in the overall shape of the collection, such as how many items are described, open to the public, and digitised. This should help future researchers explore the impact of digital access on historical research. For example, the 2016 data was analysed in Sherratt, ‘Hacking Heritage: Understanding the Limits of Online Access’, 2019.
I’ve also created a Zenodo community for National Archives of Australia historical collection data which includes this dataset, as well as annual harvests of ‘closed’ files from 2016 to 2025, and details of files digitised since 2021.
Unfortunately, due to changes in RecordSearch in 2025 it is no longer possible to harvest this sort of data.