Every VPN drops occasionally. Networks change, servers restart, mobile connections hand over between cells. The question is not whether it drops but what happens in the two seconds afterwards.
Why it disconnects
- Network handover — moving between Wi-Fi and cellular, or between access points.
- Server-side maintenance or overload, which shows up as repeated drops on one endpoint.
- Power management suspending the network adapter on laptops, especially on battery.
- Firewall or antivirus software interfering with the tunnel interface.
- Protocol mismatch with a restrictive network that blocks UDP.
The fixes, in order
Switch server first — repeated drops on one endpoint usually mean that endpoint. Then switch protocol: if UDP is being blocked, an obfuscated TCP mode will hold where WireGuard cannot connect at all. Then check power management settings on your network adapter, which is a common and easily missed cause on Windows laptops.

The part that actually matters: does the kill switch hold?
A kill switch is supposed to block all traffic the instant the tunnel fails. Our test is deliberately harsh: start a continuous ping, forcibly kill the VPN process, and count any packet that reaches the destination before the switch engages. We repeat that 40 times per provider.
Every provider in our current top ten passed — zero packets escaped across all 40 drops. That is the expected result for established providers in 2026, but it is worth verifying on your own machine, because kill switch behaviour varies by operating system and by whether the app is running with elevated permissions.
Test your own in two minutes
- Connect the VPN and start a continuous ping to a public address.
- Open your task manager or process monitor and force-quit the VPN process.
- Watch the ping. It should stop dead. If it continues even briefly, your kill switch is not engaging.
- Repeat a few times — intermittent failures are more common than total ones.
A kill switch that leaks for two seconds is worse than none at all, because you believe you are covered.
If reconnection speed is your real problem
If you move between networks constantly, the drop itself matters less than how fast it comes back. ExpressVPN's Lightway protocol re-established a usable connection in a median of 0.3 seconds after a forced network change — the fastest of anything we tested, and considerably quicker than standard OpenVPN at 3.8 seconds.
If it only happens on mobile
Mobile disconnects are usually the operating system, not the VPN. Both Android and iOS aggressively suspend background apps to save battery, and a suspended VPN client cannot maintain a tunnel. Exclude the VPN from battery optimisation and enable always-on VPN in system settings — both options exist on modern versions of each platform.
Handover between cell towers and between Wi-Fi and cellular also drops tunnels routinely. This is where reconnection speed matters more than anything else, and where WireGuard-based protocols are decisively better than OpenVPN.
If it only happens on one network
Some corporate, hotel and school networks block VPN traffic deliberately, usually by blocking UDP or inspecting SNI. The fix is obfuscation: an obfuscated OpenVPN mode over TCP 443 looks like ordinary HTTPS traffic and gets through where WireGuard cannot even connect.
Repeated drops on one network and nowhere else means the network, not your VPN. Switch to obfuscated TCP before troubleshooting anything else.



