Do VPNs Slow Down Your Internet?

Where the milliseconds actually go

Your traffic now travels to the VPN server and back before it reaches the site you asked for. If that server is in another country, you have added the round-trip time of that distance to every request — often five to forty milliseconds, and on a distant or congested route considerably more.

On top of that sits encryption, which is cheap on modern hardware and rarely the bottleneck, and the provider's own network, which is where queues form. In practice the slowest part of a VPN connection is almost never the cryptography; it is how busy the server is and how far away it sits.

Distance changes how the number feels as well as how big it is. A ten-millisecond round trip is invisible in a call and barely noticeable while browsing; sixty milliseconds on a long route puts a pause in front of every click, and you feel all of it.

What actually varies

Four variables move the number: distance to the server you picked, how many other people are on that server, the protocol in use, and how much your own line can carry in the first place.

WireGuard is the modern default and the lightest of the mainstream protocols, which is why we treat support for it as a baseline rather than a bonus. A loaded server on a long route will beat any protocol advantage you gained by connecting to it.

Density is the other half. The same city can hold a handful of servers or several hundred, and the ratio between subscribers and servers decides whether you land on a busy exit or an idle one. That is why a provider with many locations can still feel slow on the one you chose.

How much slower is realistic

A nearby server on an uncongested route typically costs you a few percent. A far server under load, or a route that leaves a continent, can cost you a third or more of a fast connection, and it will show up first in video and in large downloads.

Our speed component is an estimate built from architecture — protocol support, server density relative to subscribers, and network reach — rather than a benchmark we ran ourselves, and we say so in the methodology: How we test. Treat it as a ranking of where to look first, not as a measurement of your line.

Two measurements settle most arguments: the round-trip time to a nearby server, which the tunnel adds to every request, and the throughput your line actually carries, which the tunnel shares rather than reduces. Both take minutes to record.

What to change before you blame the VPN

Pick a server close to you, or close to the service you are using, because the total path is what matters: connecting to the country where the streaming catalogue lives beats connecting to a server three countries away.

Then check the obvious ones — a wired connection instead of wifi, a VPN protocol set to automatic or WireGuard, and a server that is not obviously the busiest one in the list. If speed is the whole reason you are here, our streaming weighting is dominated by exactly this component: Best VPN for streaming.

When a slowdown means something else

A VPN that halves your speed every time, on every server, usually points at something outside the tunnel: an overloaded exit, a rate limit in your plan, or a line that was never fast to begin with.

The cheapest way to settle it is to measure without the tunnel, then with it, on two servers in the same city. Ten minutes of that tells you more than any feature list — and it is the same discipline we apply to every number on this site: How we test.

Measuring it yourself in ten minutes

Run one test with the tunnel down and write down download, upload and ping. Connect to a server in your own country, closest available, and run it again. Then connect to a server on another continent and run it a third time.

The gap between the first two numbers is the tunnel's overhead. The gap between the second and third is distance and load, and it is almost always the larger one. If the first gap is bigger than roughly a third on a quiet nearby server, something is wrong with the route rather than with your line — that is a support question, and a fair one to put before you pay for a year.

Keep the results. Re-testing after a firmware update or a change of internet provider tells you whether a slowdown belongs to the VPN or to everything else, which is the discipline behind every number we publish: How we test.