ACLED: Coverage, Coding, Access and Citation

ACLED (Armed Conflict Location & Event Data) is an independent, non-profit conflict monitor that publishes event-level data on political violence and protest. This guide is for researchers who need to know what an ACLED record represents, how events are classified, which access routes exist, and how to cite the data and stay inside its terms. Facts below come from ACLED’s own pages, read on 2026-11-07.

What is ACLED?

ACLED describes itself as an “independent, impartial global monitor” that collects, analyses and maps data on conflict and protest. It is a registered non-profit (501(c)(3) in the United States), and its About page lists funding support from the UN Complex Risk Analytics Fund. ACLED’s attribution policy asks users to write “ACLED” or “ACLED (Armed Conflict Location & Event Data)”, keeping the ampersand, and to call it an organisation or conflict monitor rather than a “project”.

Who runs it and what it covers

According to the ACLED codebook, the data cover every country and territory in the world, with events coded from sources in over 75 languages. The geographic scope grew over time: versions 1 to 7 covered African states, version 8 added South and Southeast Asia and the Middle East, and a worldwide extension followed in 2022.

The event taxonomy: 6 event types, 25 sub-event types

The codebook organises every record under one of six event types and a sub-event type:

Event type	Sub-event types
Battles	Armed clash; Government regains territory; Non-state actor overtakes territory
Protests	Peaceful protest; Protest with intervention; Excessive force against protesters
Riots	Violent demonstration; Mob violence
Explosions/Remote violence	Air/drone strike; Shelling/artillery/missile attack; Remote explosive/landmine/IED; Suicide bomb; Grenade; Chemical weapon
Violence against civilians	Sexual violence; Attack; Abduction/forced disappearance
Strategic developments	Agreement; Arrests; Change to group/activity; Disrupted weapons use; Headquarters or base established; Looting/property destruction; Non-violent transfer of territory; Other

For research design this matters because protest and riot events sit in the same dataset as battles, so a “conflict events” count depends on which event types you filter in. Always define the filter in your methods section.

How to access ACLED data

The codebook lists three routes. Rights to use the data are conditioned on the end user licence agreement (EULA), which incorporates the content usage terms and attribution policy, so read those before downloading.

Data Export Tool. Select the countries, dates, event types and actors you need and export the result. Record the filters you chose, because the attribution policy asks you to report them.
Curated data files. ACLED publishes regional and subset files for users who want a ready-made extract rather than a custom export.
API. Programmatic access requires a free account. Sign-in links are on the ACLED website.
Read the codebook for coding rules before interpreting any variable, and note the codebook’s latest update date for your citation.
Note the access date. ACLED’s suggested dataset citation is “ACLED, accessed on [date]. www.acleddata.com.”

Citing ACLED and meeting the attribution policy

The attribution policy asks you to credit ACLED prominently, including when you have modified the data. Use a footnote with a full citation and link, or an in-text acknowledgment, and state:

the date you accessed the data;
the filters applied (countries, dates, event types, actors);
any changes or additions you made.

ACLED’s suggested forms are “ACLED, accessed on [date]. www.acleddata.com.” for the dataset and “ACLED, ‘Armed Conflict Location & Event Data (ACLED) Codebook,’ [latest update date].” for the codebook. For reports the template is Author, “Title,” ACLED, Date, URL. Visuals should carry the citation on the visual itself; social posts need a visible citation and a tag for ACLED; shared data files should include ACLED in a source column with a link. Do not use the ACLED logo without authorisation. For general practice on citing datasets, see the data citation principle.

Use restrictions that affect academic work

ACLED’s content usage terms require you to use the data responsibly, represent the data and methodology accurately in line with the codebook, state clearly how you manipulated the data, and correct inaccuracies. The terms prohibit, among other things:

harming, targeting or defaming anyone with the data;
building a competing or substitute dataset, product or platform, or benchmarking a competing product;
providing unauthorised services to others;
attributing your own analysis to ACLED;
AI, machine-learning or LLM training that creates a substitute or gives access to the data, a restriction ACLED says applies “regardless of whether such use is commercial, academic, or experimental in nature”.

ACLED may restrict or revoke access. Tier-specific access and redistribution rules sit in the EULA, not in the usage terms, so check the EULA before sharing a derived file in a repository. A data sharing agreement or a deposit in a data repository should respect those terms.

Strengths and limits

Strength: a fixed, published taxonomy. The 6 event types and 25 sub-event types are documented in the codebook, which makes coding choices inspectable and filters reproducible.
Strength: broad scope. Worldwide coverage and sources in over 75 languages allow cross-country comparison under one scheme.
Limit: an event dataset, not a count of all violence. Records are built to the codebook’s definitions, so what counts as an event, and how it is typed, is defined there. Read the codebook’s coding rules before comparing ACLED figures with another dataset.
Limit: licence conditions. Terms on competing products, derived services and AI training can constrain projects that go beyond analysis, even academic ones.
Limit: version differences. Geographic coverage was extended in stages, so confirm which versions of the codebook apply to your study period and region.

Alternatives and complements

Dataset	What its own pages say	Access and licence (as stated)
ACLED	Six event types covering political violence and protest; worldwide coverage.	Data Export Tool, curated files, API with a free account; governed by the EULA, content usage terms and attribution policy.
UCDP Georeferenced Event Dataset (GED)	Run by the Uppsala Conflict Data Program at Uppsala University; individual events of organised violence, “phenomena of lethal violence occurring at a given time and place”, geocoded to village level with single-day durations.	CSV and Excel downloads with a codebook, plus the UCDP API; the download page states a CC BY 4.0 licence.
GDELT	Monitors news media in over 100 languages; over a quarter-billion event records in over 300 categories; daily updates.	Stated as available for unlimited, unrestricted use at no fee, with citation and a link when redistributing.

The three differ in what an “event” is: UCDP’s definition is limited to lethal violence, ACLED’s includes protests and strategic developments, and GDELT draws on news media at a much larger record count. Check each codebook before merging datasets. Sources for this table: UCDP downloads and GDELT about page, read 2026-11-07. For the wider discipline context, see what is political science.

FAQ

What does ACLED stand for?

Armed Conflict Location & Event Data. ACLED asks that the ampersand be kept when the full name is used.

Is ACLED data free?

The codebook states the API requires a free account. Rights to use any ACLED data depend on accepting the EULA, content usage terms and attribution policy; tier details are in the EULA.

How do I cite ACLED?

Credit ACLED prominently, give the access date, the filters you applied and any modifications you made, and link to acleddata.com. Use ACLED’s suggested formats shown above.

Can I use ACLED data to train an AI model?

ACLED’s content usage terms prohibit AI, machine-learning or LLM training that creates a substitute or gives access to the data, whether the use is commercial, academic or experimental. Ask ACLED about any planned use.