eCFR

The Electronic Code of Federal Regulations (eCFR) is the daily-updated version of the Code of Federal Regulations (CFR), published by the Government Publishing Office. It contains the general and permanent rules established by federal agencies, organized into 50 subject-matter titles.

[!NOTE] Title 35 (Panama Canal) is reserved and contains no content. It is automatically skipped during download and conversion.

Download

Download eCFR XML files with download-ecfr. Two download sources are available.

eCFR API (Default)

The eCFR API at ecfr.gov provides daily-updated XML. This is the recommended source.

lexbuild download-ecfr --all
lexbuild download-ecfr --titles 1
lexbuild download-ecfr --titles 1-5,17

Download a point-in-time snapshot by specifying a date:

lexbuild download-ecfr --all --date 2026-01-01

The --date flag retrieves regulations as they existed on that date.

govinfo (Bulk Data)

The govinfo source provides bulk XML from govinfo.gov. This data may lag behind the eCFR API by days or weeks.

lexbuild download-ecfr --all --source govinfo

Download Options

Files are saved to ./downloads/ecfr/xml/ by default (e.g., ECFR-title1.xml, ECFR-title17.xml). Use -o to change the download directory:

lexbuild download-ecfr --all -o ./my-downloads

Convert

Convert downloaded XML files to Markdown with convert-ecfr.

Convert all downloaded titles:

lexbuild convert-ecfr --all

Convert specific titles:

lexbuild convert-ecfr --titles 1
lexbuild convert-ecfr --titles 1-5,17

Convert a single XML file by path:

lexbuild convert-ecfr ./downloads/ecfr/xml/ECFR-title1.xml

Currency Date

When converting a point-in-time snapshot, pass --currency-date to populate the frontmatter currency field with the snapshot date instead of today:

lexbuild download-ecfr --all --date 2026-01-01
lexbuild convert-ecfr --all --currency-date 2026-01-01

This makes frontmatter deterministic across runs, which is important for incremental updates. When omitted, the currency date defaults to today.

Granularity

The eCFR converter supports four granularity levels. The additional part level reflects the CFR’s part-based organizational structure.

Level	Description	Example Output
`section` (default)	One `.md` file per section	`title-01/chapter-I/part-1/section-1.1.md`
`part`	One `.md` file per part, sections inlined	`title-01/chapter-I/part-1.md`
`chapter`	One `.md` file per chapter, parts and sections inlined	`title-01/chapter-I.md`
`title`	One `.md` file per title, entire hierarchy inlined	`title-01.md`

lexbuild convert-ecfr --all -g part
lexbuild convert-ecfr --titles 17 -g title

Notes Filtering

Notes filtering works the same as for the U.S. Code:

lexbuild convert-ecfr --all --no-include-notes
lexbuild convert-ecfr --all --include-editorial-notes

Dry Run

lexbuild convert-ecfr --all --dry-run

Output Structure

At the default section granularity, output follows this directory structure:

output/ecfr/
  title-01/
    README.md
    _meta.json
    chapter-I/
      README.md
      _meta.json
      part-1/
        README.md
        _meta.json
        section-1.1.md
        section-1.2.md
      part-2/
        ...
  title-17/
    chapter-II/
      part-240/
        section-240.10b-5.md
        ...

Note that eCFR chapter numbers use Roman numerals (e.g., chapter-I, chapter-IV) while part and section numbers are Arabic. Section numbers can include dots and letters (e.g., 240.10b-5).