Every VPN adds a small performance cost. The question is never really whether a VPN slows your connection, but by how much, and whether that cost is worth what you’re getting in return. A badly configured VPN on an outdated protocol can drag a 500Mbps fibre connection down to something that struggles with a single 4K stream. A well-chosen one, on modern hardware with the right protocol, can lose you almost nothing. The only way to know which situation you’re in is to run a proper VPN speed test, rather than guessing from how fast a page feels to load.
Most guidance on this topic is written with a US broadband market in mind, which is a problem because the connection type, ISP behaviour, and server distance that shape a VPN speed test result all look different from a UK household. A test run on a full fibre connection in London behaves nothing like one run on an older FTTC line in a rural postcode, and a result taken at 9 pm on a Friday can look considerably worse than the same test at 9 am on a Tuesday, for reasons that have nothing to do with the VPN provider at all.
This guide walks through how to test correctly, what typically causes the slowdown, which protocol makes the biggest difference, and how UK broadband conditions specifically affect your results.
Table of Contents
Understanding Why a VPN Speed Test Matters
Encryption, distance to the server, and the quality of the VPN provider’s infrastructure all eat into your connection before a single web page loads. A VPN speed test is the only reliable way to see exactly how much of your bandwidth these factors are taking, rather than assuming.
What a VPN Speed Test Actually Measures
A VPN speed test compares three figures with the VPN switched off against the same three figures with it switched on: download speed, upload speed, and latency (ping). The gap between those two sets of numbers tells you the true cost of running your traffic through an encrypted tunnel rather than a direct connection.
Download speed determines how quickly data arrives, so it governs streaming quality, file transfers, and page load times. Upload speed matters less for everyday browsing but becomes important for video calls, cloud backups, and uploading large files. Latency, measured in milliseconds, is the delay before data starts moving, and it’s the figure that gaming and video calls are most sensitive to.
A VPN can leave download speed largely intact while still adding noticeable latency, so checking only one number gives you an incomplete picture. Packet loss is worth noting too, even though most consumer speed test tools don’t display it prominently; a connection losing even 1 to 2 per cent of its data packets will feel unstable in a video call or competitive game, regardless of how healthy the headline download figure looks.
Setting a Clean Baseline Before You Test
None of these figures means anything without a baseline. Before connecting to any VPN, run a speed test with it switched off, ideally over a wired Ethernet connection to remove Wi-Fi variability from the result. Close background downloads, pause any other streams in the house, and take two or three readings rather than one, since results on a home broadband connection naturally fluctuate between tests. It’s also worth restarting your router beforehand; a router that’s been running for several weeks without a reboot can quietly lose a noticeable chunk of throughput for reasons unrelated to anything you’re about to test.
How to Run a VPN Speed Test Step by Step
Testing with and without a VPN follows the same basic method: measure, connect, measure again, and compare. The difference between the two readings is your real answer, not the number either test produces in isolation.
Testing Without a VPN First
Use an independent tool such as Speedtest by Ookla or Cloudflare’s speed test rather than any speed checker built into your VPN app, since a provider’s own tool can be tuned to flatter its results. Run the test two or three times at different points in the day and note the average download speed, upload speed, and latency. This is the number every VPN reading gets compared against, so it’s worth writing the figures down somewhere rather than trusting memory once you’ve moved on to the next test.
Testing With a VPN Connected
Connect to the VPN, select a server, and wait a few seconds for the tunnel to stabilise before testing. Run the same independent speed test tool, choosing a server location on the test site that’s closer to your VPN’s exit server than to your physical location, so you’re measuring the tunnel’s performance rather than adding an extra leg of distance on top of it. Repeat this across two or three different server locations if you’re deciding between them, since a provider’s UK servers can perform very differently from its US or European ones on the same connection.
Reading Download, Upload, and Latency Results
Compare both sets of readings side by side. A drop of 10 to 20 per cent against your baseline is normal and rarely noticeable in daily use. A drop beyond 40 to 50 per cent, or a jump in latency of more than 30 to 40 milliseconds, points to a protocol, server, or provider problem worth investigating rather than something to live with. If upload speed drops far more than download speed, that’s often a sign the VPN server itself is under heavy load from other users rather than a fault with your own connection.
The Protocol Effect on VPN Speed Test Results
The single biggest factor in your VPN speed test result is which protocol your app is using, not which provider you’ve subscribed to. Two people on the same VPN service can see very different numbers purely because one is running WireGuard and the other is stuck on an older default.
Why Protocol Choice Changes Your Numbers
WireGuard and OpenVPN handle encryption in fundamentally different ways. WireGuard runs in the operating system’s kernel with a small, modern codebase of around 4,000 lines, which eliminates much of the processing overhead that slows older user-space protocols. OpenVPN, by contrast, has to hand every packet between kernel and user space, and its codebase runs to tens of thousands of lines supporting a much wider range of configurable cyphers.
Independent benchmark testing on a 1Gbps connection has recorded WireGuard sustaining roughly 940 to 960 Mbps, compared with around 480 Mbps for OpenVPN over UDP, with OpenVPN over TCP performing worse still due to how TCP handles retransmissions within another TCP connection, sometimes called the TCP-over-TCP problem. Separate data overhead testing puts WireGuard’s protocol overhead at around 4.5 per cent, against roughly 17 to 20 per cent for OpenVPN, depending on whether it’s running over UDP or TCP.
Connection setup speed follows the same pattern. WireGuard typically establishes a working tunnel in under a second, while OpenVPN’s handshake can take five to eight seconds, which is noticeable if you’re reconnecting frequently on a mobile connection or switching between Wi-Fi and mobile data throughout the day.
Which Protocol to Pick for Speed
For most home users in 2026, WireGuard (or a provider’s rebranded version of it, such as NordLynx) is the sensible default and typically retains 90 to 95 per cent of your base connection speed. OpenVPN still has a place if your network blocks UDP traffic or you need to disguise VPN traffic as ordinary web browsing over TCP port 443, but expect 20 to 40 per cent slower speeds as the trade-off for that flexibility.
IKEv2/IPsec sits in between and is worth choosing on mobile devices for its ability to survive switching between Wi-Fi and mobile data without dropping the tunnel, which matters more for reliability than for raw throughput. Newer protocol implementations that move OpenVPN’s data channel into the kernel are narrowing the gap on some platforms, but WireGuard still leads on most measures as things stand. Our guide to setting up a VPN securely covers how to switch protocols inside the most common apps.
Typical VPN Speed Test Results by Protocol (1Gbps Connection)
| Protocol | Typical Throughput | Speed Retained | Connection Setup |
|---|---|---|---|
| WireGuard | ~940-960 Mbps | 90-95% | Under 1 second |
| OpenVPN (UDP) | ~480 Mbps | 60-80% | 5-8 seconds |
| OpenVPN (TCP) | ~320 Mbps | 50-65% | 5-8 seconds |
| IKEv2/IPsec | ~600-700 Mbps | 75-85% | 1-2 seconds |
UK Broadband Conditions That Affect Your VPN Speed Test
Your VPN speed test result depends as much on your UK broadband connection as on the VPN itself, and this is the part that most testing guides written for a US audience skip entirely.
Full fibre (FTTP) now reaches 82 per cent of UK homes and gigabit-capable broadband is available to 89 per cent (27.1 million homes), according to Ofcom’s Spring 2026 Connected Nations update. That matters for VPN testing because the connection type you’re on sets the ceiling for what a VPN can possibly retain.
A VPN cutting 15 per cent off a 900Mbps full fibre line still leaves you with a fast connection; the same percentage cut on an older FTTC or ADSL line, still in use across a meaningful share of UK premises, can push you close to buffering territory. If your VPN speed test results look disappointing, it’s worth checking what your underlying connection type actually is before assuming the VPN is to blame.
Peak-Time Congestion on UK Networks
Many UK providers manage network traffic more actively in the evening, broadly between 7pm and 11pm, when network demand peaks. A VPN speed test run during this window can look considerably worse than the same test at midday, and the difference is often the broadband connection queuing behind other households rather than the VPN itself.
Testing at two different times of day, and comparing both against a same-time baseline, separates ISP congestion from genuine VPN overhead. Our testing on UK streaming services found this pattern repeatedly when benchmarking protocols on a residential connection in Manchester, with results shifting noticeably once the evening peak began.
Server Distance and Latency for UK Users
Choosing a VPN server physically close to you, ideally London or Manchester for UK users, keeps the added latency from a VPN speed test to single- or low-double-digit milliseconds. Connecting to a server in another continent for no particular reason adds distance-related latency on top of the protocol overhead, which stacks two separate slowdowns together and makes the VPN look worse than it actually is.
This distance effect matters even more for latency-sensitive uses; our guide to VPNs and online gaming breaks down how UK ISP routing and server choice affect ping specifically, rather than just download speed, including how routing through the London Internet Exchange can shorten the path for UK-to-Europe connections.
How to Improve a Slow VPN Speed Test Result
If your VPN speed test shows a larger drop than the ranges above, the fix is usually one of a small number of things rather than a fundamental limitation of using a VPN.
Switch Protocol or Server Location
Start with the simplest change first. If your app is set to OpenVPN by default, switching to WireGuard in the settings menu is often the single biggest improvement available, frequently recovering 20 to 40 per cent of the speed lost to protocol overhead alone. If you’re already on WireGuard, try a different, closer server location before assuming the provider itself is at fault, and check whether the server you’ve picked appears heavily loaded in the app, since a crowded server behaves like a slow one regardless of protocol.
Businesses managing VPNs for remote staff should also weigh protocol choice against network requirements rather than speed alone; our guide to firewalls and VPNs covers this from an organisational perspective, including where OpenVPN’s flexibility still earns its place despite the speed cost.
Fix Router and Device-Level Issues
Restart your router and modem before testing, since a connection that’s been running for weeks can accumulate minor performance drift unrelated to the VPN. Connect via Ethernet rather than Wi-Fi wherever possible to remove wireless interference from the result, and close other bandwidth-heavy applications, particularly background updates, before running a test.
Older or budget routers can also struggle to handle VPN encryption at higher speeds if you’ve set up a VPN at the router level rather than on an individual device; this is worth checking if every device on your network shows the same slowdown, since the bottleneck may be the router’s processor rather than the VPN connection itself.
Digital nomads and frequent travellers face an additional variable, since protocol performance and legality shift by country and network; our guide to choosing a VPN for travel sets out which protocols perform best in which regions, and where an obfuscated or TCP-based connection becomes necessary despite the speed cost.
Testing on mobile networks introduces its own quirks worth knowing about before you assume something is wrong. A VPN speed test on 5G will naturally show more variation between attempts than one on fixed-line broadband, since mobile signal strength changes as you move and as the surrounding cell tower load shifts through the day.
Mobile processors also handle encryption less efficiently than desktop or router hardware, so the same protocol can show a slightly larger percentage drop on a phone than it does on a laptop connected to the same network. Running a mobile VPN speed test at the same time and location on different days gives a more honest picture than a single reading taken while walking between rooms or moving through an area with a patchy signal.
Frequently Asked Questions
These are the questions that come up most often once people have run their first VPN speed test and are trying to work out whether their numbers are normal.
What is a VPN speed test?
A VPN speed test measures download speed, upload speed, and latency both with and without a VPN connected, so you can see exactly how much performance the VPN’s encryption and routing are costing you.
Does a VPN always slow down internet speed?
Almost always, yes, though the amount varies hugely. A modern protocol like WireGuard on a nearby server can cost as little as 5 to 10 per cent of your base speed, while an older protocol on a distant, crowded server can cost well over half. The rare exceptions are connections where an ISP is actively throttling certain types of traffic, such as gaming or streaming data; in those cases, a VPN can occasionally appear faster than no VPN at all, simply because the encrypted tunnel prevents the ISP from recognising and deprioritising that traffic.
What counts as a good VPN speed test result?
A drop of 10 to 20 per cent against your baseline is normal and rarely noticeable. Anything beyond a 40 to 50 per cent drop, or added latency beyond 30 to 40 milliseconds, is worth investigating rather than accepting.
Why is my VPN speed test worse in the evening?
UK broadband providers tend to manage traffic more actively during peak evening hours, broadly 7 pm to 11 pm, when overall network demand is highest. This ISP-side congestion can look identical to a VPN problem in a speed test, so it’s worth testing at a different time of day before concluding the VPN itself is at fault.
Is there a free VPN speed test tool?
Yes. Independent tools such as Speedtest by Ookla and Cloudflare’s speed test are free and don’t require a VPN subscription, and using an independent tool rather than a provider’s built-in tester avoids any risk of a result being tuned to look favourable.
Which VPN protocol is fastest for UK users?
WireGuard, in almost all cases. It consistently retains the highest percentage of base connection speed among widely available protocols, and most major UK-facing VPN providers now offer it as either the default or an option.