You connect to a VPN, check your public IP address, and see a different one. Yet a familiar website still remembers your account or preferences. Is the VPN failing? Not necessarily. An IP address is only one signal websites can use. Login sessions, cookies, and browser characteristics can remain unchanged when your connection takes a different route.
To understand why websites still recognise you when using a VPN, test these mechanisms separately. The following checks help you see what changed without mistaking a privacy score for proof of anonymity.

Start With a Baseline
Before recording any results, choose your VPN setup and install any necessary software. A browser extension may handle only supported browser traffic, while a device-level VPN generally covers more applications, depending on its configuration and split-tunnelling rules.
For this experiment, a free VPN connection can provide an alternative network route. Complete the setup first, but leave the VPN disconnected until your baseline measurements are finished.
Now use your usual browser and internet connection. Open an IP-checking website and record your public IP address, approximate location, and network provider. IP-based locations can be inaccurate; they do not establish your precise whereabouts or identity.
Next, run the Electronic Frontier Foundation’s Cover Your Tracks test. It checks certain tracking protections and compares your browser’s characteristics with configurations in its observed dataset. Its results do not represent every browser on the internet.
Record the attributes shown, such as language, time zone, screen details, and fingerprinting results. Also note whether you are signed into the familiar website you plan to revisit.
Keep your browser configuration unchanged throughout the comparison. With these baseline results recorded, you are ready to test the VPN connection.
Change the Connection, Then Compare
Connect the VPN without clearing cookies, signing out, or changing browsers. The important measurement is what your browser actually reports.
- Refresh the same IP-checking website.
- Record the new public IP address and reported location.
- Repeat Cover Your Tracks in the same browser mode.
- Revisit the familiar website and note what it remembers.
- Disconnect the VPN and repeat the IP check.
When tested traffic passes through the VPN, websites normally see the VPN server’s public IP rather than your original address. However, this does not automatically remove cookies, end account sessions or change browser characteristics.
Compare individual fingerprint attributes rather than just the overall score. Results can vary between runs, including when browsers randomise certain values. A different score alone does not show that the VPN caused the change.
An unchanged fingerprint also does not demonstrate failed VPN encryption. These tests examine different mechanisms.
If you want to repeat the experiment with a device-level VPN on Windows, check the setup instructions for the Windows VPN application. After changing your VPN configuration, establish a fresh baseline before running another comparison.
Separate Account Recognition From Stored Website Data
A website displaying your account name while you are signed in needs no fingerprinting explanation: an active login session already connects your visit to that account.
Cookies and other website storage can also preserve preferences or identifiers. To investigate their role, keep your network connection and VPN state fixed while making these observations:
- In a regular window, visit the website while signed in. Record the personalised details shown.
- Sign out and revisit in that same regular window. Record what remains.
- Close all existing private windows. Open a fresh private browsing session, visit the same page without signing in, and compare again.
These stages separate signing out from switching browser modes. If account details disappear after signing out, the active session likely explained them. If other preferences disappear only in the fresh private session, stored website data may have contributed.
Treat this as evidence, not a definitive diagnosis. Private browsing can also change tracking protections and extension behaviour. It does not guarantee anonymity, and a website showing no personalisation could still collect identifying signals.
Understand What a Fingerprint Can Tell You
Browser fingerprinting combines characteristics exposed by your browser and device. Mozilla’s fingerprinting documentation explains how details such as fonts, screen information and other attributes can contribute to identification, and how Firefox limits some of them.
A distinctive test result does not prove that a particular shop or social platform used fingerprinting to recognise you. The test demonstrates a potential identification method; it does not inspect that website’s internal decision-making.
Use this table to keep the conclusions proportionate:
Observation | Reasonable interpretation | What it does not prove |
| Public IP changes | Tested traffic has a different public-facing route | Complete anonymity |
| Fingerprint attributes stay similar | Browser characteristics remain consistent | VPN encryption failed |
| Account name remains visible | An active login session may identify you | Your original IP leaked |
| Simulated tracking requests are blocked | Those test requests were blocked | All tracking is prevented |
| IP-based location changes | The new address maps to another location | GPS or location permissions changed |
Check Leak Warnings in Context
DNS and WebRTC tests answer additional questions, but neither explains every case of website recognition.
A DNS test reports resolvers observed handling its queries. Compare them with your intended VPN, browser and network settings. An unfamiliar resolver is not automatically a leak: browser-configured encrypted DNS or a third-party service may explain it. Equally, encryption alone does not establish that DNS followed the intended VPN route.
For WebRTC results, distinguish private local addresses from your original public IP. A private address alone does not establish that your public IP leaked. If the test exposes your original public address while the VPN is connected, investigate routing, split tunnelling and browser settings.
Repeat unexpected results before drawing conclusions. Passing these checks does not prevent identification through accounts, cookies or fingerprints.
Act on the Results
Match each change to the issue you found. Review browser tracking protection and website permissions. Remove unnecessary extensions, especially those with broad access to browsing activity.
Separate browser profiles can keep account sessions and some stored data apart, but they do not guarantee different fingerprints. Avoid adding numerous privacy extensions without understanding their effects; unusual configurations can sometimes become more distinctive.
Keep a simple before-and-after record: connection state, public IP, browser mode, login status and relevant test attributes. That record is more useful than an isolated anonymity score.
Changing your IP changes one part of what websites see. Testing accounts, storage and browser characteristics separately gives you a clearer basis for deciding what to change next.


















