Choosing sustainable file formats is essential for keeping your data usable, shareable, and secure over the long term. The more you understand how open or closed your current file formats are, the better you can plan for long-term access.

As you shift from saving your data as you work on it, to saving for long-term preservation, something to consider is your file formats. The format your data is saved in is one of the most critical elements in how reusable your data will be in the future. 

This information should be documented in your project README file, lab notebook, and/or Data Management and Sharing (DMS) Plan.

1. Understand your data

Consider:

  • Data security – Do formats support encryption, access controls, or redaction?
  • Data storage – How large are the files, and how will that scale over time?
  • Data sharing & reuse – Will collaborators and future users be able to open and understand your files?

2. Identify data types and formats

Types

List the types of data you will collect or create:

  • Tabular data
  • Survey data
  • Experimental measurements
  • Models and software
  • Audiovisual data (audio, video)
  • Geospatial data
  • Physical samples (with associated digital records)

Formats

For each type, note the file format(s):

  • Plain text (.txt)
  • Comma-separated values (.csv)
  • Geo-referenced TIFF (.tif, .tfw)
  • Sequencing (FASTQ)
  • Tissue Slide Images (.tiff, .svs)

3. Use open and widely accepted formats

Prefer formats that are:

  • Open – Not tied to a single vendor or proprietary software
  • Well-documented – With stable, public technical specifications
  • Widely used – Supported by multiple tools and a large user community

Open formats:

  • Can be used by a variety of proprietary, free, and open-source tools
  • Are more likely to remain readable even if the original software becomes unavailable
  • Often have publicly available documentation, which supports long-term preservation and reuse

Using open, well-documented, and widely adopted formats:

  • Improves the long-term usability and integrity of your data
  • Makes it easier to share and archive data in repositories
  • Increases confidence in your preservation strategy

If you need to convert or migrate your data to another format, be aware of potential data loss or corruption. Test conversions where possible and retain original files until you have verified the results.

Below are commonly used, preservation-friendly formats (where appropriate). Always confirm any specific repository or disciplinary guidance:

  • Databases: XML, CSV
  • Geospatial: SHP, DBF, GeoTIFF, NetCDF
  • Moving images (video): MOV, MPEG, AVI, MXF
  • Audio: WAVE, AIFF, MP3, MXF
  • Numeric / statistical data: ASCII, DTA (Stata), POR (SPSS portable), SAS, SAV (SPSS)
  • Images: TIFF, JPEG 2000, PDF, PNG, GIF, BMP, OME-TIFF
  • Text: PDF/A, HTML, ASCII, XML, UTF-8–encoded text
  • Graphs / structured configuration data: JSON, YAML, XML

4. Estimate amount and size of data

Estimate:

  • Total volume of data you expect to produce over the life of the project
  • Rate of data production (e.g., per day, per experiment, per instrument run) and whether it may change

These estimates help you plan for storage, backup, and long-term preservation needs.

The Fred Hutch Data Science Lab provides useful information about file sizes for types of data commonly generated in biomedical research: Data File Size Details

5. Check repository and community guidance

Many repositories and organizations provide explicit recommendations or preferred formats. For example:

The Library of Congress also provides guidance for long-term digital preservation:

Consult these resources, along with practices in your discipline and among similar organizations, to select formats that balance usability, accessibility, and preservation over time.