26 June 2026 Ancello

A Genealogy File Naming System That Keeps Digital Records Organized

Use a consistent genealogy file name that sorts well, preserves identity and source clues, and still works outside your current software.

A useful genealogy filename should tell you enough to recognise a record without becoming a miniature citation. Put the most sortable information first, use a controlled vocabulary, add a person or household identifier, and keep the complete source description in your research system rather than forcing every detail into the file name.

Use one predictable pattern

A practical starting pattern is:

date_record-type_surname-given_place_repository-reference.ext

For example:

1911-04-02_census_quayle-margaret_douglas_uk-census_rg14-34672.webp

1898-10-12_marriage_cain-eliza_quayle-william_douglas_iom-registry_42-18.pdf

about-1925_photo_cain-family_peel_private-album_p03.tif

This pattern sorts chronologically, groups similar records when searched, and remains readable if the file is copied outside the workspace.

Put dates in ISO order

Use YYYY-MM-DD, YYYY-MM, or YYYY. Year-first dates sort correctly on most systems and avoid ambiguity between day/month and month/day conventions. Preserve uncertainty explicitly:

about-1880
before-1914
between-1901-and-1911
undated

Do not turn an approximate year into 1 January merely to fill a pattern. The filename should never imply more precision than the source.

Standardise record types

Choose a short list and keep it in a README: birth, baptism, marriage, death, burial, census, directory, probate, military, immigration, newspaper, letter, photo, and oral-history.

Avoid switching between marriage-cert, wedding, and mar for the same category. Add a subtype only when it improves discovery, such as newspaper-obituary or military-service-record.

Name the subject consistently

Use surname-given for one person and add a second subject when the event requires it. For a household census, the head or a stable family identifier can be the subject, while metadata links the file to everyone on the page.

Names change and spellings vary. Pick a filing form for the filename, but record the original spelling in the source transcription. For married women, a birth surname is often more stable for filing; add the spouse or married name when needed to distinguish the person.

Same-name people need a disambiguator such as birth year, place, or workspace ID:

1901_census_kelly-john-b1872-peel_douglas_...

Do not put sensitive identifiers for living people in filenames that may be synchronised or shared.

Keep place names concise

Use the place needed to distinguish the record, not a full historical hierarchy. A short standard form such as douglas-iom or liverpool-lancs works well. Preserve the record’s exact jurisdiction in metadata and the citation, especially when boundaries changed.

Add a source clue, not the whole citation

A repository abbreviation and record reference helps prevent collisions and locate the record if it is separated from the workspace. Keep a controlled list of abbreviations. Do not use a session URL or a 150-character catalogue title in the filename.

For downloaded records, capture the full collection, database, page, image, URL, and access date in the citation. The filename is a label; the citation is the route back.

Use safe characters

For portability across Windows, macOS, cloud storage, scripts, and websites:

use lowercase letters, numbers, hyphens, and underscores;
avoid slashes, colons, question marks, quotes, and trailing periods;
avoid leading spaces and very long paths;
use a normal hyphen instead of typographic punctuation;
keep the extension accurate.

Lowercase reduces accidental duplicates on systems that treat case differently.

Handle multi-page and edited files

Add ordered page numbers with leading zeros: p01, p02, p03. If pages form one logical document, retain the page images and optionally create a combined PDF.

Use suffixes that describe purpose, not repeated “final” versions:

_master for an untouched preservation scan;
_access for a smaller sharing copy;
_transcription for a text version;
_annotated for a marked working copy.

Prefer version history in managed storage. If versions must be filenames, use a date or zero-padded number and document which one is current.

Roll out the system safely

Inventory file types and existing conventions.
Draft the pattern and controlled vocabulary.
Test it on 30 varied files.
Check sort order and search on every device in use.
Write a one-page README with examples.
Rename in small batches while keeping a backup and mapping old to new names.
Update links in your genealogy software or workspace.

Bulk renaming without a mapping log can break thousands of media links. Work on copies first and verify a representative sample after every batch.

What belongs outside the filename

Do not overload names with every person on a census, a full archive citation, copyright conditions, a long description, transcription, confidence analysis, or private notes. Store those as metadata and links. A short stable filename works because the workspace supplies the rich context.

A five-second test

Show a filename to another relative. Can they tell approximately when the item was created, what it is, who or what it concerns, and how to distinguish it from a similar file? If yes—and the complete citation remains connected—the system is doing its job.