VPN and Latency: Does a VPN Increase Ping? (Explained)
A VPN sends your traffic through an extra server, so latency almost always rises: the added distance (your location to the VPN server and from there to the destination) plus encryption and server load. A nearby server on a modern protocol such as WireGuard adds only a few milliseconds. In rare cases a VPN can lower ping by taking a better route, but that is the exception.
On this page
Quick answer: a VPN (virtual private network) encrypts your traffic and sends it through a server run by the VPN provider. That detour makes the path longer, and every extra kilometre of fibre costs about 1 ms of round-trip time per 100 km. So ping with a VPN is usually higher than without. How much higher depends on how far away the VPN server is, how busy it is, and which protocol is used. A close server on WireGuard adds a few milliseconds; a far server can add a hundred or more.
How a VPN changes the path
Without a VPN, your packets go from you to the destination along the route your ISP chooses. With a VPN:
- Your device encrypts the packet and sends it to the VPN server.
- The VPN server decrypts it and forwards it to the destination from its own IP address.
- The reply returns to the VPN server, is encrypted again and comes back to you.
Total latency ≈ latency(you to VPN server) + processing + latency(VPN server to destination) + the same on the way back. If the VPN server is in the same city as you, the first leg is tiny and the second may be nearly your normal route. If it is on another continent, you pay for that distance twice (out and back) even when the destination is next door. See undersea cables and the speed of light.
Where VPN latency comes from
| Component | Typical effect |
|---|---|
| Extra distance (detour) | Usually the biggest factor; about 1 ms of RTT per 100 km of fibre path |
| Encryption and decryption | Small on modern hardware (well under a few ms), larger on weak routers or phones |
| Server load and bandwidth | Queueing on a crowded server adds variable delay and jitter |
| Protocol overhead | Packet headers reduce usable packet size; TCP-based tunnels can add stalls |
| Peering of the VPN provider | Good transit can make the second leg faster than your ISP's route; poor transit makes it slower |
| DNS through the VPN | May use the provider's resolver, which can be quicker or slower than yours |
Protocols and their latency behaviour
- WireGuard (and protocols based on it): lean code, UDP-based, generally the lowest overhead.
- IKEv2/IPsec: efficient, handles network changes well (good on mobile).
- OpenVPN over UDP: flexible, a bit heavier than WireGuard.
- OpenVPN over TCP: useful to pass restrictive firewalls but can suffer "TCP over TCP" problems: if packets are lost, both layers retransmit, causing large delays. Avoid it for real-time use if possible.
Background on the transport layers is in TCP and UDP explained.
Can a VPN ever make ping better?
Yes, occasionally. The internet does not always choose the shortest path; routing depends on commercial agreements and policies (see routing, BGP and peering). If your ISP reaches a game server through a congested or roundabout route, a VPN provider with better peering might reach it via a faster path. Gamers sometimes see this effect. It is hit and miss: the only way to know is to measure with and without the VPN, at the time of day you play.
VPN and gaming
For online games, every added millisecond and every spike matters. Use a VPN for gaming only if you need it (privacy on public Wi-Fi, a blocked service, a DDoS risk) or if testing shows it improves your route. Choose a server close to you, or one close to the game server; avoid free or overloaded servers. See latency for gaming.
VPN and work (corporate VPN)
A full-tunnel corporate VPN sends all traffic to the office or a cloud gateway, so a cloud app that is geographically close to you can become slow if your traffic is sent via a distant head office. Split tunnelling, which sends only corporate traffic through the VPN, or regional gateways, often fixes this. If video calls are choppy over a corporate VPN, ask IT whether call traffic can bypass it. See video calls and VoIP.
How to test your VPN's effect
- Disconnect the VPN and run a GlobeLatency test with a fixed selection and 10 passes. Save the report link.
- Connect to the VPN server you intend to use. Wait a few seconds for it to settle.
- Run the same selection and number of passes again.
- Compare overall average, best and worst, and a few key destinations. A difference in best values reveals the added distance; a difference in worst values reveals instability.
- Repeat with another server, and at another time of day.
Note that a VPN changes your apparent location. The "You" marker on the map and any geolocation will reflect the VPN exit, and the world comparison continues to compare the same destinations, so percentages will mostly show the effect of the detour. See Accuracy and limitations.
Worked example
You are in Sofia and test a European game server: 38 ms without a VPN. With a VPN server in Frankfurt (about 1,300 km away) the result is 62 ms: roughly +24 ms. With a VPN server in New York the result is around 170 ms: your traffic crosses the Atlantic to the VPN server and the VPN server's request crosses it again to reach a game server that was only a short hop away, and the replies make the same trip back. If your destination were in the US, however, that same New York server would sit close to it, and the penalty would shrink or even disappear.
Tips to keep VPN latency low
- Use the nearest server; check the provider's own ping figures if displayed.
- Prefer WireGuard or UDP-based modes.
- Avoid overloaded servers at peak hours; try another in the same city.
- Use a wired connection; Wi-Fi plus VPN stacks two sources of variation.
- Enable split tunnelling if you only need the VPN for some apps.
- Keep the VPN app and device firmware current.
Common mistakes and myths
- "A VPN lowers ping." Rarely. Usually it raises it.
- "VPN speed equals VPN latency." A fast download through a VPN can still have high latency.
- "Any server in my country is fast." Large countries still add distance, and some servers are overloaded.
- Testing the VPN once at a quiet hour and assuming it is always the same.
What to do next
Read how to reduce latency, or follow troubleshoot high latency to confirm whether a VPN is your cause. Then run a before and after test.
Frequently asked questions
Does a VPN increase ping?
In most cases yes. Traffic takes a detour through the VPN server, so the path is longer, and encryption adds a little processing. A nearby, well-run server on a fast protocol may add only 2 to 10 ms, while a distant or overloaded server can add 100 ms or more.
Can a VPN reduce latency?
Sometimes. If your ISP has a poor route to a destination, the VPN provider may have a better path, and the detour can end up faster. It is uncommon, and you must verify it by measuring with and without the VPN.
Which VPN protocol has the lowest latency?
WireGuard-based protocols generally add less overhead than OpenVPN, especially OpenVPN over TCP. IKEv2 is also efficient. UDP-based modes are preferable to TCP-in-TCP tunnelling, which can compound delays when packets are lost.
How should I choose a VPN server for gaming?
Pick a server physically close to you, or close to the game server if your own route is poor, then measure. Ping both to the VPN server and from it to the game. Avoid crowded servers during peak hours.
How do I measure the effect of my VPN on latency?
Run the same test with the same destinations and passes with the VPN off and on, and compare averages, best and worst. GlobeLatency makes this easy, but remember that a VPN also changes the apparent location of your IP address.