Clear standards.
Accountable records.

How each case is sourced, classified, written, verified, and transformed into public data.

One source file

data/cases.csv is the main source file. Corrections begin there so every case page, chart, map, and downloadable file can be rebuilt from the same reviewed information.

cases.csv
case pages JSON GeoJSON statistics

Download data

Download the main CSV or one of the generated files for research, mapping, analysis, or future integrations.

Source standards

Records prioritize dedicated missing-person profiles, government and law-enforcement bulletins, official missing-child posters, public alerts, and established case-information organizations. Every profile connects readers to an active primary and secondary source.

Narrative standards

Case writing is neutral and factual. Confirmed events, official statements, investigative theories, and unresolved questions remain distinct. The database does not present a theory as fact or state that a crime occurred unless the cited record supports that classification.

Location data

Coordinates represent generalized city-level last-seen locations in WGS 84 decimal degrees. They support geographic context without implying an exact disappearance point. State boundaries come from the U.S. Census Bureau's 2025 TIGERweb States layer and are stored locally with the website.

Quality review

  1. Confirm identity, missing date, age, and last-seen locality.
  2. Verify both source links and reconcile conflicting information.
  3. Review classifications and language for unsupported conclusions.
  4. Validate timelines, images, coordinates, and generated outputs.
  5. Update the verification date only after completing the review.

Limitations

Public information can be incomplete, inconsistent, delayed, or removed. Media prominence can change. Absence from this database does not indicate that a case is less important or no longer active.

Read the complete methodology document