How to test my internet speed accurately (2026)
Quick Answer
To test your internet speed accurately, first remove as many local bottlenecks as possible, then run multiple tests under consistent conditions and compare the results with your service plan. The most reliable method is to test on a computer connected to your router by Ethernet, with other devices and background downloads paused. If Wi‑Fi is what you want to assess, test that separately after you have established a wired baseline.
Overview
Internet speed tests are only useful if you control the conditions. A result can be limited by your device, Wi‑Fi signal, router load, background syncing, VPN use, or the test server itself, so a single quick test on a busy wireless network does not tell you much. The best approach is to test in stages: first confirm your broadband connection with a wired device, then test Wi‑Fi performance in the places where you actually use it. Run more than one test and note the time of day, because congestion can vary. Focus on the main figures the test reports: download speed, upload speed and latency. Download affects streaming and general browsing, upload matters for video calls and cloud backups, and latency affects responsiveness in gaming, calls and remote work. A good result is not just a high number; it is a consistent result that makes sense for your package, device and connection method. If your wired results are poor, the issue is more likely with the internet service, router or line. If wired results are good but Wi‑Fi results are poor, the problem is usually wireless coverage, interference or device placement.
Who this is for
Home users, renters, homeowners and small office users who want a dependable picture of broadband and Wi‑Fi performance before contacting their provider, moving equipment or upgrading service.
What you’ll need
- A computer or laptop, ideally with an Ethernet port or Ethernet adapter
- An Ethernet cable if you want the most accurate baseline test
- Access to your router
- A recognised speed test service or your broadband provider’s own test
- A notepad or phone notes app to record results, times and test conditions
Before you start
Check whether you want to test the internet connection itself or your Wi‑Fi experience. For the internet connection, use a wired connection if possible. Pause large downloads, cloud backups, software updates, streaming and game updates on all devices you can control. Turn off any VPN unless you specifically want to measure VPN performance. If possible, restart the device you will test from if it has been heavily used, and make sure your web browser and operating system are behaving normally.
Step-by-step
- 1
Decide what you are measuring
Choose whether you want a baseline broadband test, a Wi‑Fi test, or both. If you want to know whether your provider is delivering the service properly, start with a wired test from a computer connected directly to your router. If you want to know whether your phone or laptop works well in a certain room, do a separate Wi‑Fi test there afterwards.
Why: This prevents confusion. A poor Wi‑Fi result does not automatically mean your internet service is slow, and a good wired result can help you narrow the problem to wireless coverage rather than the broadband line.
- 2
Prepare the network and test device
Stop or pause bandwidth-heavy activity on the device you are testing from and, if possible, on other devices in the home or office. Close streaming apps, cloud sync tools, online backups and large downloads. Disconnect or disable a VPN unless you are testing VPN speed. If you can, connect your computer to the router with Ethernet and leave Wi‑Fi off on that computer during the wired test.
Why: Speed tests measure what is available at that moment. Other traffic on the network, or traffic routed through a VPN, can reduce the result and make it harder to tell whether the connection itself is the issue.
- 3
Run a wired baseline test first
Use a recognised speed test in a web browser or app and run the test on the wired computer. Let the test complete fully so you can see download, upload and latency. Run more than one test, spaced a little apart, and record the results along with the time of day. If the service lets you choose a test server, use an automatic or nearby server unless you have a specific reason to test a distant one.
Why: A wired baseline removes most Wi‑Fi variables. Repeating the test helps you spot whether a poor result is a one-off or a pattern, and recording the time can reveal peak-time congestion.
- 4
Check whether the wired result makes sense
Compare the wired results with the speed information for your broadband package and with the provider’s own guidance on how speeds are measured. Look for consistency across repeated tests rather than focusing on a single best result. If your speeds are much lower than expected, repeat the test later and, if practical, test with a second device or another Ethernet cable.
Why: Service plans, device limits and line conditions all affect results. A consistent shortfall on wired tests is stronger evidence of a service or equipment issue than one disappointing test.
- 5
Test Wi‑Fi separately in real use locations
After the wired test, switch to Wi‑Fi and test in the rooms where performance matters, such as a home office or living room. Stay still during each test and test one location at a time. If your device can use different Wi‑Fi bands and your router exposes them separately, note which one you are connected to. Record the location, signal strength if shown, and each result.
Why: Wi‑Fi speed depends heavily on distance, walls, interference and band choice. Separate room-by-room testing shows whether the issue is coverage in a specific area rather than a whole-home broadband problem.
- 6
Repeat at different times
Run the same tests again at a quieter time and at a busier time, such as when more people are likely to be online. Keep the method the same each time: same device, same location, same connection type if possible.
Why: Internet performance can vary by time of day because of local or provider network congestion. Consistent testing conditions make comparisons meaningful.
- 7
Interpret the pattern and act on it
If wired speeds are good but Wi‑Fi is poor, focus on router placement, interference, mesh or extender setup, or device limitations. If wired and Wi‑Fi are both poor across repeated tests, restart the router, check for provider outages or line issues, and contact your provider with your recorded results. If only one device is slow, troubleshoot that device rather than the whole connection.
Why: Patterns point to the right fix. Without separating wired and wireless results, it is easy to spend money or time on the wrong problem.
Why this works
An accurate speed test isolates variables. Ethernet reduces interference and shared-airwave issues, repeated tests reduce the effect of one-off fluctuations, and comparing wired and Wi‑Fi results helps distinguish broadband issues from local wireless problems. Recording time, device and location turns a vague impression of slowness into evidence you can act on.
Common mistakes to avoid
- Testing over Wi‑Fi first and assuming the result reflects the broadband line
- Running a test while updates, backups or streaming are active elsewhere on the network
- Using only one test result and treating it as definitive
- Leaving a VPN on without realising it can affect routing and speed
- Testing from an old or overloaded device that may be the bottleneck
Troubleshooting
Wired speed is much lower than expected
Repeat the test later, restart the router, try a different Ethernet cable or device, and check your provider’s service status page. If the problem persists across repeated wired tests, contact your provider with your recorded results.
Wired speed is fine but Wi‑Fi is slow in some rooms
Move the router to a more open and central position if possible, reduce obstacles and interference, and consider mesh Wi‑Fi or a better access point layout if coverage remains poor.
Results vary wildly from one test to the next
Make sure nothing else is using the connection heavily, test at the same time with the same device and method, and try a second recognised speed test service to rule out a server-specific issue.
One device is slow but others are fine
Update that device, restart it, forget and reconnect to Wi‑Fi if relevant, and check whether the device itself has limited network capability or background activity causing the slowdown.
The speed test will not complete or stalls
Try another browser or the provider’s own test, disable browser extensions that interfere with web traffic, and check whether a firewall, VPN or captive portal is affecting the test.
Compare your options
Wired Ethernet test
Best for: Checking the actual broadband connection with the fewest local variables
Pros: Most reliable baseline, avoids Wi‑Fi interference, easier to compare with your service plan
Cons: Needs a suitable device and cable, does not reflect the experience on wireless devices
Wi‑Fi test near the router
Best for: Checking whether Wi‑Fi performance is broadly healthy when signal is strong
Pros: Quick, convenient, useful for spotting whether wireless can perform well at all
Cons: Can still vary with device quality and nearby interference
Wi‑Fi test in everyday rooms
Best for: Assessing real-life performance where you actually work or stream
Pros: Shows coverage problems and dead spots, helps with router placement decisions
Cons: Less suitable for judging the provider connection itself
Provider’s own speed test
Best for: Support cases and package-related checks
Pros: May align better with provider diagnostics and support processes
Cons: May not be directly comparable with third-party tests
| Option | Best for | Pros | Cons |
|---|---|---|---|
| Wired Ethernet test | Checking the actual broadband connection with the fewest local variables | Most reliable baseline, avoids Wi‑Fi interference, easier to compare with your service plan | Needs a suitable device and cable, does not reflect the experience on wireless devices |
| Wi‑Fi test near the router | Checking whether Wi‑Fi performance is broadly healthy when signal is strong | Quick, convenient, useful for spotting whether wireless can perform well at all | Can still vary with device quality and nearby interference |
| Wi‑Fi test in everyday rooms | Assessing real-life performance where you actually work or stream | Shows coverage problems and dead spots, helps with router placement decisions | Less suitable for judging the provider connection itself |
| Provider’s own speed test | Support cases and package-related checks | May align better with provider diagnostics and support processes | May not be directly comparable with third-party tests |
Alternatives
- Use your broadband provider’s diagnostic tool if you need results that support a fault report
- Use a router app or Wi‑Fi analyser to assess signal strength and coverage, not just speed
- Test with a second device to confirm whether the problem is device-specific
Pro tips
- Record device, connection type, room and time of day with each result
- If you are comparing changes, alter one thing at a time, such as router position or test device
- Use the same test service when comparing before-and-after results
- For home working and video calls, pay attention to latency and stability, not only download speed
Safety notes
- Route Ethernet cables safely to avoid trips or strain on ports
- Do not reset or reconfigure your router unless you know your provider login details or setup method
- Avoid opening provider equipment unless the manufacturer documentation says it is user-serviceable
What this guide does not cover: This guide covers how to test internet speed accurately, not how to optimise advanced router settings, redesign a network, or diagnose faults inside provider-owned infrastructure.
Cost considerations
Accurate testing can help you avoid unnecessary upgrades. If wired performance is already in line with your package, spending on a faster plan may not solve poor Wi‑Fi in specific rooms; better placement or wireless equipment may be the more effective cost.
Frequently asked questions
Should I test on Wi‑Fi or Ethernet?+
Use Ethernet first if you want the most accurate view of the broadband connection itself. Use Wi‑Fi afterwards if you want to measure the experience on wireless devices in real rooms.
How many times should I run a speed test?+
Run more than once and compare the pattern. Repeating the test under the same conditions helps you tell the difference between a one-off fluctuation and a consistent issue.
Why is my phone slower than my laptop?+
Different devices have different wireless capabilities, software load and background activity. If one device is consistently slower while others are fine, the device is often the limiting factor.
Does a VPN affect a speed test?+
Yes, it can. A VPN changes how your traffic is routed and can reduce speed or increase latency, so turn it off unless you specifically want to test VPN performance.
What matters more: download, upload or latency?+
It depends on what you do. Download matters for most browsing and streaming, upload matters for backups and video calls, and latency matters for responsiveness, gaming and voice or video quality.
The basic testing method changes slowly, but provider tools, app interfaces and consumer Wi‑Fi technology can change over time.