How to Use OONI Explorer to Track Internet Censorship Worldwide

Updated

OONI Explorer is the public interface to the Open Observatory of Network Interference (OONI) dataset — described on ooni.org as "the world's largest open dataset on internet censorship," with millions of measurements collected from more than 200 countries since 2012. You use it when you want evidence about whether a specific site or app is blocked, throttled, or reachable in a given country, rather than a general claim that "the internet is censored there." It is a browsing and analysis tool, not a live test of your own connection — for that you run OONI Probe, which feeds the same dataset.

What OONI Explorer actually contains

Each record in Explorer comes from an OONI Probe measurement: a test run on a real network somewhere in the world. The dataset is built from two main test families referenced on the site:

  • Website and app blocking tests — check whether specific sites or services are reachable, and classify anomalies (for example, signs of DNS tampering, TCP/IP blocking, or HTTP interference).
  • Network performance tests — measure speed and performance of the network the probe ran on.

Because measurements are contributed by volunteers, coverage is uneven: some countries and time periods have dense data, others are sparse. Explorer shows you what exists, not a guaranteed sample of every network in a country.

How to look up censorship in a country or for a specific site

The workflow below matches how the dataset is organized. Exact button labels may change as the site is updated, so treat the steps as the sequence of filters rather than fixed pixel positions.

  1. Open OONI Explorer from ooni.org (the "Explore OONI Data" entry point).
  2. Choose your filter dimension. You can browse by:
    • Country — pick the country you care about.
    • Time range — narrow to the period you're investigating (useful for tying results to an event such as an election or protest).
    • Test type / measurement — restrict to website blocking, app blocking, or performance, depending on your question.
  3. Read the results as measurements, not verdicts. A single anomalous measurement is a signal, not proof. Look for a pattern: the same site failing across many probes, on multiple networks, over a sustained window.
  4. Open individual measurements to see the detail — which test ran, what the network returned, and how the result was classified (reachable, anomaly, confirmed block, etc.).
  5. Cross-check against OONI's own reporting. The site's Research and Blog sections publish analyses that interpret the raw data, for example country-specific findings on blocked categories of sites or throttling during protests. Comparing your filter results with a published analysis helps you avoid misreading an isolated anomaly.

Reading the results without over-claiming

The most common mistake is treating one red result as a confirmed nationwide block. Keep these distinctions in mind:

What you see What it likely means What to do next
Isolated anomaly for one site Could be a local network issue, a probe error, or a real block Check whether other probes in the same country/time show the same site failing
Many anomalies, same site, same country, same period Strong signal of blocking Look for OONI's published analysis or corroborating reports
Consistent anomalies across many sites/categories Possible broad filtering or throttling Compare against the time range to see when it started
No data for a country/period Under-measurement, not "no censorship" Widen the time range or check whether probes exist there

Absence of measurements is not evidence of a free internet. It usually means too few volunteers ran OONI Probe on that network during that window.

Using findings for research, reporting, or advocacy

Explorer's value for a write-up or campaign comes from being specific and reproducible:

  • Cite the scope: country, time range, test type, and number of measurements you looked at.
  • Show the pattern, not one dot: reference multiple measurements rather than a single anomalous result.
  • Link to the raw measurement so others can verify it.
  • Pair data with context: OONI's blog analyses show how measurement data is combined with interviews and events (for example, work on censorship of specific services or throttling during protests). Following that model keeps your claim defensible.

If your goal is to add data rather than read it, install and run OONI Probe — its measurements become part of the same public dataset you're browsing in Explorer.