The Complete Guide to Archival Cataloguing

Cataloguing is the work that makes an archive real. A collection that is not catalogued may exist physically, but it cannot be searched, cited, shared, or properly protected. A catalogue is not a bureaucratic overhead — it is the condition for everything else the archive does.

This guide covers the essentials: what archival cataloguing is, which standards apply, how to structure a catalogue record, how to tackle a backlog, and how AI-assisted tools are changing what is practically achievable for archives of any size.

---

What is archival cataloguing?

Archival cataloguing is the process of creating structured, descriptive records for the items in a collection. Each catalogue record captures the essential facts about an item — what it is, who created it, when, where it came from, and who can access it — in a consistent, searchable format.

A good catalogue serves three audiences at once. For the researcher, it is an archival finding aid: the means by which they identify relevant material and navigate the archive collection. For the institution, it is a management tool for tracking holdings, monitoring condition, and demonstrating stewardship to funders. For future archivists, it is the record of decisions made.

Cataloguing is distinct from related activities. Digitisation converts physical items to digital files — it does not automatically create catalogue entries. Indexing creates access points for retrieval; it is a component of cataloguing, not a replacement for it.

---

Core principles of archival description

Three principles underpin all good cataloguing practice.

Respect des fonds. Maintain the original groupings and original order of a creator’s records. Do not intermix records from different creators or impose an artificial logical order. Provenance is meaningful data — a catalogue that destroys it is less useful than the uncatalogued collection it replaced.

Multilevel description. Most archive collections are hierarchical: a fonds contains series; series contain files; files contain items. Good cataloguing describes at the appropriate level, capturing contextual information once rather than repeating it at item level descriptions across hundreds of records. This is the central principle of ISAD(G) and the basis of professional archival description in the United Kingdom.

Consistency over perfection. A catalogue where every item has a reference number, brief description, date range, and access conditions is searchable. A catalogue where 20% of items have beautifully detailed catalogue entries and 80% have nothing is not. Apply the same level of detail uniformly — especially when clearing a backlog.

---

Cataloguing standards and guidance

UK archives and museums work across several established standards. The right choice depends on your context. For detailed guidance on choosing between them, see our guide to archival cataloguing standards.

ISAD(G) is the international standard for archival description, published by the International Council on Archives. It defines 26 descriptive elements across seven areas and supports multilevel description. Most UK archive management systems are built around it. A revised international standard — Records in Contexts (RiC) — is in development and will eventually replace ISAD(G), though ISAD(G) remains current best practice.

SPECTRUM is the UK collections management standard for museums. It defines 21 collection management procedures and is the basis for Arts Council England Accreditation. It is procedural rather than a field-level data model — it tells you what processes to follow, not what fields to complete.

Dublin Core is a lightweight 15-element standard useful for digital collections, community archives, and situations where full ISAD(G) is disproportionate to the archive’s resources.

EAD (Encoded Archival Description) is the XML encoding standard for archival finding aids. It structures your catalogue data for machine exchange with national aggregators such as Archives Hub, funders, and partner institutions. For a full explanation, see our guide to the EAD metadata standard.

---

What goes in a catalogue record?

Regardless of standard, every good catalogue record contains:

• Reference code — a unique reference number used for physical labelling and cross-referencing; permanent once assigned
• Title — a brief, descriptive label, either taken from the item or constructed by the archivist
• Date(s) — date of creation or accumulation, as precise as the evidence allows
• Level of description — fonds, series, file, or item
• Creator — the person, family, or organisation that created or accumulated the records
• Scope and content — what is in the collection: record types, subject matter, date range, significant absences
• Extent — volume of material (number of items, files, or linear metres)
• Access conditions — restrictions, embargoes, personal information considerations

The catalogue structure at each level of description should reflect the original groupings of the material, not an artificial arrangement imposed by the cataloguing archivist.

---

Cataloguing specific collection types

Different materials have different cataloguing requirements. Use these guides for your specific collection type:

• Photograph collections require format, photographic process, condition, and copyright fields that general catalogue records do not. See our guide to cataloguing a photograph collection.
• Museum collections follow object-centred conventions — object name from a controlled vocabulary, maker, materials, dimensions, and current location. See our guide to cataloguing a museum collection.
• Making your catalogue discoverable is the ultimate goal. Our guide to making an archive searchable online covers the practical steps from metadata to public access. And for context on why cataloguing backlogs matter, see the hidden cost of uncatalogued collections. If your project involves funded digitisation, our guide to NLHF metadata standards including ISAD, EAD3, and Dublin Core explains what funders expect.

---

Tackling a cataloguing backlog

Most archives have a backlog. Some practical approaches:

Survey before you start. Walk the collection at a high level and produce a brief survey record — approximate volume, date range, record types, condition. This surfaces surprises before they derail the cataloguing work.

Set a minimum viable catalogue record. For backlog-clearance work, agree in advance what the minimum acceptable record contains: reference code, title, date, brief description, extent, access conditions. That minimum is the target.

Batch similar material. A group of 200 letters from the same correspondent shares vocabulary, reference sources, and provenance context. Organise the work to exploit that — it is significantly faster than cataloguing 200 unrelated items.

AI-assisted cataloguing. For digitised collections, AI tools can cut the time to a described, searchable catalogue dramatically. The Archiver reads each digitised file — extracting text via OCR, identifying document types, pulling out dates, names, and places — and generates a draft catalogue record. You review and approve; the AI handles the data entry. It exports in EAD3, BagIt, and CSV so your catalogue entries are ready to share with funders or aggregators from day one. Request early access to try it on your own collection.

For a detailed comparison of approaches — spreadsheets, manual cataloguing, CMS platforms, and AI-assisted tools — see our comparison page.

---

Start cataloguing

The catalogue you have been putting off is more tractable than it looks. The standards are clear, the guidance is available, and the tools now exist to make the data entry manageable. See our features overview for a full picture of what AI-assisted cataloguing can do, explore the dedicated page for archivists for profession-specific workflows, or learn how it works step by step.

Request early access to try The Archiver on your own collection.