Wi-Fi vs Ethernet Speed Test: Why Wireless Results Read Lower
Click Below To Share & Ask AI to Summarize This Article
Save time and get the key takeaways instantly. Choose your favorite AI assistant to read and analyze this page for you.
Pay for 500 Mbps, test on your phone in the living room, get 180. Plug a laptop into the router, get 470. Nothing is wrong with your broadband in that scenario. The speed test is measuring your whole path, and Wi-Fi is usually the narrowest part of it.
This guide shows how big the Wi-Fi gap is in real UK homes using our own test results, explains how our speed test estimates whether you're wired or wireless, and sets out a simple method for comparing the two fairly, so you know whether to call your provider or fix your Wi-Fi.
Data as of October 2026
Figures are medians from 5,911 fixed-line tests run on BroadbandSpeedTest.uk.com from UK connections between 1 March and 15 September 2026 (Starlink and mobile networks excluded, as are VPNs and data centres). See our methodology and the latest monthly speed report.
The Short Answer
A wired test measures your broadband line. A Wi-Fi test measures your line plus your router's radio, the walls in between, your neighbours' networks and your device's wireless chip. In our data, tests we classed as likely wired had a median download of 651 Mbps; tests classed as likely Wi-Fi had a median of 142 Mbps. Part of that gap is because people who test over a cable tend to have faster packages, but even on the same group of networks the difference is large, as the tables below show.
If you want to know whether you're getting the speed you pay for, test with a cable. If you want to know what your devices actually experience, test over Wi-Fi in the rooms you use. Both are useful; they answer different questions.
Wired vs Wi-Fi: What Our Tests Show
Of 5,911 fixed-line tests, our test labelled 73% as likely Wi-Fi, 20% as mixed and only 6% as likely wired. Most people test on whatever device is in their hand, which is one reason crowd-sourced speed figures always sit below the speeds providers advertise.
| Connection (estimated) | Tests | Download | Upload | Ping | Jitter |
|---|---|---|---|---|---|
| Likely wired | 376 | 651 Mbps | 110 Mbps | 10 ms | 1.3 ms |
| Mixed / unknown | 1,208 | 319 Mbps | 94 Mbps | 32 ms | 3.8 ms |
| Likely Wi-Fi | 4,327 | 142 Mbps | 63 Mbps | 42 ms | 14 ms |
To strip out some of the "faster package" effect, here is the same split within network groups. Customers of the same group of providers have broadly similar line types, so the gap here is closer to the real cost of Wi-Fi:
| Network group | Likely wired | Mixed | Likely Wi-Fi |
|---|---|---|---|
| Independent full fibre (altnets) | 701 Mbps (85) | 536 Mbps (235) | 385 Mbps (715) |
| Openreach-based ISPs (BT, Sky, TalkTalk, Zen and others) | 538 Mbps (173) | 145 Mbps (610) | 84 Mbps (2,352) |
| Virgin Media | 758 Mbps (29)* | 374 Mbps (160) | 374 Mbps (595) |
Median download, with number of tests in brackets. *Fewer than 30 tests, so treat as indicative.
On altnet full fibre, where almost everyone has a fast line, likely Wi-Fi tests came in at about 55% of the likely wired median. On Openreach-based providers the gap is much wider. That's mostly because the wired testers there skew towards full-fibre customers while the Wi-Fi testers include many on part fibre, which is a reminder that these labels are estimates and the groups aren't perfectly matched. Your own before-and-after comparison, done the way described below, is far more reliable than any average.
How Our Test Spots Wi-Fi vs Wired
Web browsers aren't allowed to tell a website whether you're on Wi-Fi or Ethernet, so we estimate it from jitter: how much your ping varies from one measurement to the next while the connection is otherwise idle. Wi-Fi shares airtime between devices and has to retry when signals collide, which makes response times wobble. A cable doesn't. After each test we apply three simple bands:
- Under 2 ms jitter: labelled "Likely Wired"
- 2–5 ms: labelled "Mixed / Unknown"
- 5 ms or more: labelled "Likely Wi-Fi"
It's a rule of thumb rather than a reading of your device settings. A congested line can push a wired connection over 5 ms, and a strong Wi-Fi 6 connection right next to the router can sneak under 2 ms. Here's how download speed and ping change across finer jitter bands in our data:
| Jitter | Tests | Median download | Median ping |
|---|---|---|---|
| Under 1 ms | 97 | 773 Mbps | 8.6 ms |
| 1–2 ms | 279 | 613 Mbps | 11 ms |
| 2–5 ms | 1,203 | 320 Mbps | 32 ms |
| 5–10 ms | 1,505 | 234 Mbps | 37 ms |
| 10–20 ms | 1,245 | 146 Mbps | 42 ms |
| 20 ms or more | 1,571 | 93 Mbps | 52 ms |
The pattern is steady: every step up in jitter comes with lower download speeds and higher ping. Some of that is Wi-Fi, some is congestion on slower lines, and they often go together. Either way, a high jitter reading on your result is a strong hint to repeat the test over a cable. Our ping and jitter explainer covers why jitter matters for calls and gaming.
Why Wi-Fi Tests Read Lower
- Walls and distance. UK homes are hard on Wi-Fi: solid brick, plasterboard with foil backing, chimney breasts and kitchen appliances all absorb 5 GHz signals. Speeds can halve from one room to the next.
- Shared airtime. Wi-Fi is a shared medium. Every phone, TV, doorbell and smart speaker takes turns, and the slowest devices take the longest turns.
- Neighbours. In flats and terraces, dozens of networks can overlap on the same channels, especially on 2.4 GHz.
- Your device. Many phones and older laptops have two antennas and older Wi-Fi standards. An older Wi-Fi 5 laptop may never exceed a few hundred Mbps however close it is to the router. See what Wi-Fi 6 changes.
- Upload suffers more. Your device transmits with far less power than the router, so upload falls off faster than download at a distance.
Cables have limits too. A standard gigabit Ethernet port tops out at roughly 940 Mbps of real throughput, so on packages above 1 Gbps you need a 2.5 Gbps port and adapter to see the full speed. Old Cat5 cables, 100 Mbps USB adapters and some powerline adapters will cap you well below that.
How to Run a Fair Wi-Fi vs Ethernet Test
The point is to change one thing only, the connection, and keep everything else the same.
- Use one device. A laptop with an Ethernet port (or a decent USB-C adapter) is ideal. Testing a wired laptop against a wireless phone tells you about the phone, not the Wi-Fi.
- Quieten the house. Pause streaming, game downloads and cloud backups on other devices, and close other tabs and apps.
- Wired first. Plug directly into a router port, not via a switch or powerline adapter. Run three tests and note the middle download and upload.
- Wi-Fi in the same room. Unplug, connect to the 5 GHz (or 6 GHz) network within a couple of metres of the router, and run three more tests.
- Wi-Fi where you actually use it. Repeat three tests from the room where things feel slow.
- Compare middle results. Note our "Likely Wired" or "Likely Wi-Fi" label and the jitter figure each time.
Run the whole sequence within about 15 minutes so time-of-day effects don't muddy the comparison (see broadband speed by time of day). For a full checklist, read how to test your broadband speed accurately.
What Your Comparison Tells You
Wired is close to your package, Wi-Fi is much lower everywhere
Your line is fine. The router's Wi-Fi or your device is the limit. Try 5 GHz, a better channel or a newer router. Start with our Wi-Fi speed tips.
Wi-Fi is fine near the router but collapses in other rooms
That's a coverage problem. A mesh system or a wired access point in the problem area usually fixes it; a faster package won't.
Wired is also well below your package
The problem is the line, the router or the network. Wired results are the evidence your provider will accept. Work through why is my broadband slow, then see how to complain about broadband speed.
You can compare your wired result with other customers of your provider on its page, for example BT, Sky or Virgin Media, but remember those figures are mostly Wi-Fi tests too.
FAQ
Is Ethernet always faster than Wi-Fi?
Nearly always in practice, and it's more consistent. A modern Wi-Fi 6 or 7 device right next to a good router can match a gigabit Ethernet port on download, but Ethernet still wins on ping and jitter, which is what matters for gaming and calls.
My test says "Likely Wi-Fi" but I'm plugged in. Why?
The label comes from jitter, not your device settings. Jitter of 5 ms or more on a cable usually means something else is loading the connection, such as another device streaming or downloading, or congestion on the line or network. Re-test with the house quiet. Powerline adapters also add jitter and often show as mixed or Wi-Fi.
Does Wi-Fi affect ping?
Yes, usually by a few milliseconds on a good connection and much more on a weak or busy one. Wi-Fi mainly adds variation (jitter), which is why online games feel worse over wireless even when average ping looks fine. Our gaming ping guide has target figures.
Which result should I use when complaining to my provider?
A wired result. Providers measure speeds at the router and will normally ask you to test over a cable before they treat slow speeds as a line fault. Keep the Wi-Fi results too: they show what you actually experience, and help you decide whether a better router is worth it.