What is browser fingerprinting? How it works and what it reveals

Check your own browser fingerprint

The quickest way to understand fingerprinting is to see your own. The free browser fingerprint test reads the same signals described below and shows the visitor ID they produce.

Run the free browser fingerprint test →

How browser fingerprinting works

Every browser answers questions any page may ask: what time zone is it, which languages, how big is the screen, how many CPU cores, which fonts are installed. Other answers come from asking the browser to do work. Drawing text and shapes on a hidden canvas or a WebGL scene produces slightly different pixels on different graphics cards, drivers and operating systems; processing a sound through the Web Audio API produces slightly different numbers.

A fingerprinting script collects dozens of these answers. None of them is unique alone, but together they usually single out one browser on one device.

Signals a browser fingerprint uses

• Canvas fingerprinting: the exact pixels of text and shapes drawn on a hidden canvas.
• WebGL: the GPU vendor and renderer strings, supported extensions and a rendered 3D image.
• Audio: the output of a short sound processed through the Web Audio API.
• Fonts and emoji: which fonts are installed and how emoji are drawn.
• Environment: screen size, colour depth, time zone, languages, platform, CPU cores and memory.
• Capabilities: speech voices, media codecs, WebRTC and WebGPU support, CSS features.

What browser fingerprinting is used for

Advertising networks have used fingerprints to follow people across sites, which is why browsers now limit some signals. The other main use is security: banks, shops and SaaS products recognise a returning device to stop account takeover, fake sign-ups, repeat free trials, coupon abuse and referral fraud. Used this way, the fingerprint stays on the service that collected it and is not shared for advertising.

Is browser fingerprinting accurate?

A fingerprint changes when the browser updates, a font is installed or the window moves to another screen. Hashing all signals into one value makes any change produce a new ID. Robust systems match each signal separately on the server and accept a returning browser when most of them still agree, weighting rare values more than common ones.

Identical devices are the hard case: two phones of the same model and software version look almost the same. A security system should treat such a shared device class with extra checks rather than merge the two people.

Anti-fingerprinting browsers

Brave randomises several signals per site, Firefox can resist fingerprinting by reporting standard values, and the Tor Browser makes all its users look alike. These measures reduce cross-site tracking. They also make a browser stand out to a security system, which can then ask for a human check instead of trusting the request outright.

See your own browser fingerprint

The free FingerPrint Checker browser fingerprint test shows the visitor ID your browser gets, whether it survives a private window, and which signals lowered its trust score.

Run the free browser fingerprint test →

Frequently asked questions