Skip to content

English version, machine-translated from Spanish. The Spanish original is the one that counts.

How we document each case

Detrás de cada punto del mapa hay una persona. Esta es la forma en que reunimos, verificamos, nombramos y contamos sus historias, hoy 3,267 registros públicos en 32 Estados, explicada paso a paso.

1. What we document

We record violence against lesbian, gay, bisexual, transsexual, transgender, travesti, intersex, queer, asexual, non-binary and other (LGBTTTIQANB+) people that occurred in Mexico, and the memory of those who were part of the community. Each record is a person, not an event or a press article.

Record typeWhat it includes
Crime or offenseMurders and other documented attacks, with or without a complaint filed with a prosecutor’s office.
DisappearanceMissing or unlocated persons, including those later found alive or dead.
ObituaryProfiles of community members who died, even if no violence was involved: activists, artists, friends and figures who are part of our history.
OtherCases that do not fit the previous categories, explained in their record.

Además, cada caso indica la situación de la persona: falleció, falleció tras una agresión, desaparecida / no localizada, con vida / sobreviviente.

The project began recording in 2025, but the documentation traces stories back to 1530. That the records span so many years does not mean we have followed them since then: we reconstruct them from the traces that remained.

2. Where the data comes from

This is a community record, not an official count. The figures are the trace of the voices left along the way, gathered from public sources:

  • Press: articles from national, local and sexual-diversity outlets.
  • Verified social media: posts by families, friends, collectives and organizations, cross-checked with other sources.
  • Community archives: obituaries and memorials the community itself has built over the years, which we take up with their authorship, date and link.
  • Search notices and alerts: bulletins from search commissions, Amber Alert and Protocolo Alba.
  • Citizen reports: what families, friends and organizations send with the form to report a case, with your details or anonymously. An anonymous report needs at least one link to a public source so it can be verified.
  • Documentation teams: collectives from different states that share this platform and the same care protocols. Each case indicates which team documented it (meet the sources).

3. How we verify

No case is published automatically. A report arrives as pending, moves to under review and ends as approved o not applicable; whoever reports with their details can check the progress with their tracking number.

  • Cross-checking sources. We look for more than one source to confirm the person’s name, date, place and connection to the community.
  • Only what is explicit. Information about identity or orientation is recorded only if a source says it about that person. We do not infer it from context (for example, from the fact that they worked in a bar or took part in a pageant) or from rumors.
  • Cases with weak documentation. If a case has only one source or the person’s identity is unclear, we publish it and it counts in the statistics, with a note on the documentation at the end of their account. We do it this way because in the most distant years coverage of these deaths is scarce or nonexistent, and excluding them would erase them again.
  • Duplicates. Before creating a record we search for the person by name, pseudonym, place, date and source link. If two records turn out to be the same person, they are merged into one.
  • Sources in plain sight. Each record lists its public sources, so anyone can review them.

4. Name and identity

Chosen name, always

We publish the name by which the person identified and was known. The first name changes; surnames are kept. The legal name, if different, is kept privately for the internal record and never appears on the site, in the open data or in the search; if a source uses it, we omit it or remove it from the public text. When we do not yet know the social name, the person appears as “"N"” followed by their surnames.

Gender identity, orientation and expression

We record each dimension separately and respecting the person’s self-definition, without pathologizing, in accordance with the Supreme Court Protocol and the Yogyakarta Principles:

Gender identity
Mujer Transgénero, Hombre Transgénero, Mujer Cisgénero, Hombre Cisgénero, No Binario, Muxe, Travesti, Otra.
Sexual orientation
Lesbiana, Gay, Bisexual, Pansexual, Asexual, Heterosexual, Diversidad sexual (sin especificar), Otra.
Gender expression
Femenina, Masculina, Andrógina, No conformista, Genderqueer, Otra.

Muxe is a gender identity of the Zapotec culture of the Isthmus of Tehuantepec, in Oaxaca. We record it as a separate identity because that is how these people name themselves, even though there is no official registry of this identity in Mexico.

When there is no information, the data is left as “Unknown”. It is not the same as “no”: it only means that no source mentions it. Pronouns (she, he, they) are recorded if known.

Intersectionality

Violence does not weigh equally on everyone. That is why we also record whether the person was a migrant, spoke an Indigenous language, had a disability, was homeless or defended human rights, and their ethnic background. As with identity, we only mark what a source documents.

5. Date and place

  • Main date: that of death; if unknown, that of the events or the disappearance; and failing that, that of the discovery. It is the one used by the map, the timeline and the charts by year.
  • Place: that of the events or of the discovery, with the names and codes of states and municipalities from INEGI’s Geostatistical Framework. In obituaries it is the place of death.
  • Approximate coordinates: the published ones are rounded to about 110 meters, so as not to reveal homes.
  • If only the municipality is known, the map shows a reference point and, when zooming in, lights up the entire municipality with the note “Only the municipality is known, not the exact place”. If only the state is known, the record lights up the state. We never invent an exact point, and downloads do not carry coordinates that do not exist.

6. How we count

  • One person, one record. If there were several victims in the same event, each one has their own record.
  • An attack is not a death. When the person survived a crime or attack, in the statistics it counts as attack (survived), not alongside the people who were killed, and their record says so clearly.
  • Disappearances. They count as a disappearance even if the person was found, and we distinguish those still missing, those who appeared alive and those found dead.
  • Crimes. They are the classification documented in the sources or the investigation file, not a sentence. A case can have several.
  • Age. The age at the time of the events, grouped in ranges; if unknown, it counts as “No data”.
  • Public cases only. The figures for Open data, the API and the downloads are calculated only with reviewed and published cases.
  • Tasa por millón de habitantes. En el Report dividimos las muertes violentas documentadas entre la población de cada año y la expresamos por cada millón de habitantes al año. La población de cada año la estimamos con el Censo 2020 y la Encuesta Intercensal 2025 del INEGI, suponiendo un crecimiento constante entre los dos conteos. Sin año elegido, la tasa es el promedio de los últimos cinco años completos.
  • Rango probable y cifras inestables. Junto a cada tasa damos un rango probable: un intervalo de confianza del 95 % que considera el azar. Con menos de 10 casos marcamos la tasa como «cifra inestable».
  • Homicidios en general, como contexto. Junto a la tasa de cada Estado mostramos todos los homicidios que registra el INEGI con los certificados de defunción, por cada 100 mil habitantes. Sirven para ver si un lugar es violento para toda la población. No dicen cuántas de esas personas eran LGBTTTIQANB+, porque ningún registro oficial lo anota.
  • Tasa experimental sobre la población LGBTI+ y trans (solo para todo el país). Dividimos los casos documentados de personas de 15 años o más, o sin edad conocida, entre la población que contó la Encuesta Nacional sobre Diversidad Sexual y de Género (ENDISEG) 2021 del INEGI: 5 millones de personas LGBTI+ y 908.6 mil personas trans y de otras identidades de género, de 15 años y más. Para la tasa trans contamos a las personas trans, travestis y muxes. La llamamos experimental porque la encuesta se hizo una sola vez, cuenta solo a quien se reconoce LGBTI+ y quiso decirlo, y tiene su propio margen de error, que nuestro rango no suma. Por Estado no la calculamos porque ese margen es muy grande.

Example

A trans woman was attacked in 2019 and survived; in 2021 a gay person disappeared and months later was found dead.

The first counts in 2019 as attack, not as a death. The second counts as disappearance found dead, with the date of death if known. Both add to the total of documented people.

7. What we publish and what we do not

PublicPrivate (only the documenting team)
Social name, photograph (optional), age, identity, date, state and municipality, approximate location, summary and account based on public sources, sources and profile. Legal name, investigation file and criminal case numbers, health data, detailed causes of death, family contacts and details of the person reporting.

No revictimization. Causes of death and the context of death are used only in aggregate in the statistics and do not appear on each person’s record. We do not publish images of the violence, and photos are stored without metadata. We ask the same of anyone who uses these data: non-sensationalist language and naming each person according to their identity.

8. What we do not measure

  • It is not an exhaustive record. Much violence is not reported, is recorded under the legal name and the gender assigned at birth, or never reaches the media. The real figures are higher.
  • The geography also reflects our capacity. That a state has more cases may be because there is more press, more collectives or a documentation team there, not only because there is more violence.
  • Distant years are incomplete. Coverage of these deaths in past decades is scarce; comparisons between years should be read with care.
  • We do not judge the motive. Recording a case does not assert that it was a hate crime nor does it replace a prosecutor’s investigation: it documents that a member of the community was a victim of violence and that their memory matters.
  • No sabemos cuántas personas LGBTTTIQANB+ hay en cada lugar ni cada año. La única cifra oficial es la ENDISEG 2021, de todo el país. Por eso nuestra tasa principal se mide contra toda la población, y la que se mide contra la población LGBTI+ y trans es experimental.
  • These are not official figures. They may differ from those of prosecutors’ offices, INEGI or other observatories, which use other sources and other criteria.

9. Corrections and removal

Records are corrected when new information appears: each important change is noted in the case’s internal follow-up, with the previous value. Any person, family or organization can tell us about an error, provide a source or ask us to remove or reduce a record from the correction link on each page, which reaches us with the page indicated.

Report an error Report a case

10. Reference framework

Documents this methodology is based on:

  • Supreme Court of Justice of the Nation (2022). Protocol for adjudicating with a perspective of sexual orientation, gender identity and expression, and sex characteristics. General Directorate of Human Rights, first edition.

    Basis of the concepts we use to record sexual orientation, gender identity and expression, and of the language with which we name each person.

    See it at the SCJN (opens in another tab)

  • Yogyakarta Principles (2006) and Yogyakarta Principles plus 10 (2017). Principles on the application of international human rights law in relation to sexual orientation, gender identity, gender expression and sex characteristics.

    They support the self-definition of identity, the rejection of pathologization and the right to memory and truth of sexually and gender-diverse people.

    See the principles (opens in another tab)

  • National Institute of Statistics and Geography (INEGI). Geostatistical Framework: catalog of states and municipalities, with their codes and boundaries.

    Names, codes and boundaries of the states and municipalities with which we locate each case on the map.

    See it at INEGI (opens in another tab)

  • Instituto Nacional de Estadística y Geografía (INEGI). Censo de Población y Vivienda 2020 y Encuesta Intercensal 2025.

    La población de cada Estado y de cada año, con la que calculamos las tasas.

    See it at INEGI (opens in another tab)

  • Instituto Nacional de Estadística y Geografía (INEGI). Estadística de Defunciones Registradas: defunciones por homicidio.

    Los homicidios de toda la población, que mostramos como contexto junto a cada Estado.

    See it at INEGI (opens in another tab)

  • Instituto Nacional de Estadística y Geografía (INEGI). Encuesta Nacional sobre Diversidad Sexual y de Género (ENDISEG) 2021.

    La población LGBTI+ y trans de 15 años y más, con la que calculamos la tasa experimental del país.

    See it at INEGI (opens in another tab)

The data are published under the CC BY 4.0 license. To use them in your research or report, see the open data guide, with the dictionary of each field, how to cite us and the API.