Every name needs a source.
Sources checked November 8, 2026
This memorial brings together public records of people reported killed during or in connection with the 2025–2026 Iran protests, including subsequent protest-related executions. It is an evolving memorial, not a complete or independently established death toll.
The source lists
| Source | How it is used |
|---|---|
| HRANA — The Crimson Winter | Confirmed, named protester, child and civilian entries in Appendix A. Unconfirmed entries, unnamed groups and state personnel are excluded. Forty later protest-related execution records from the October annual report and November 3 and 8 reports were also compared. |
| Iran International — Javidnaman | 3,431 published memorial identities and source-reported numerical ages. Burial addresses are not interpreted as locations of death. |
| Hengaw — January 2026 memorial database | 1,214 primary profiles across all 51 listing pages. The database identifies its scope as the January 2026 crackdown. A review of 135 profile narratives supplied explicit ages and locations and distinguished repeated profiles from different people sharing a name. |
| Iran Human Rights — September 2026 report | 3,895 verified-annex records were compared to identities already established in the other in-scope lists. The annex includes state-affiliated deaths without a category for every row, so unmatched identities are held for review. The report's correction identifying Mohammad Mahan Raei as alive overrides older death lists. |
| IranVictims community dataset | 5,482 community entries were used for source-ID crosswalks and existing English/Persian spellings where identity is consistent with the primary source. Mismatched translations and unsafe crosswalk rows were rejected. The community dataset is never the sole evidence for a death or an age. |
Identity matching
Exact source identifiers and consistent names link records first. Exact normalized names may then be matched with supporting dates, provinces or ages. Longer exact names across different sources can be linked when there is no conflicting information. Transliteration variants are compared conservatively. A shared short name alone does not establish identity. Explicit primary-source aliases and reviewed duplicate profiles can establish a match; known distinct same-name profiles remain separate.
Unresolved homonyms or spelling variants may remain as separate entries. The 6,974 visible entries must not be read as a count of distinct people killed. Each entry retains its primary source identities so corrections can be traced.
Ages and uncertainty
An age is displayed only when a documenting source explicitly reports a positive numerical age. Conflicting ages are left blank, with the disagreement explained in the source card. No ages are inferred from photographs, the word “child,” birth dates, community estimates or missing values. Conflicting reported death dates are also omitted. Source-provided partial names are labeled as partial identities.
Known-alive corrections, government-role conflicts, unnamed groups, and unresolved scope or identity matches are withheld. Inclusion reflects the cited source's reporting; Golabii has not independently investigated each case. These are public names and source references only.
The original draft
The supplied design contained 2,600 generated placeholder names. Those test records have been removed entirely. Blank carvings below the names symbolize those whose identities are not available; they do not represent fabricated people or a numerical estimate.
Inspect the data
Download CSV · Download JSON with source provenance
For additions or corrections, contact the documenting organization linked in the person's record. Published source corrections should be reflected before the next update.