Everyone knows Ethernet is faster than Wi-Fi. But knowing something and actually measuring the difference are two separate things. I have run this test in dozens of configurations over the years, and the results consistently surprise people — not because Ethernet is better, which everyone knows, but because Wi-Fi is so much worse than people assume, especially under realistic conditions.
The Numbers: A Direct Comparison
Here are measured averages from a typical residential setup: a modern Wi-Fi 6 (802.11ax) router, a laptop sitting about 8 meters away with one wall between it and the router, versus a direct Gigabit Ethernet connection on the same network. All tests run during off-peak hours to isolate the medium, not congestion:
| Metric | Gigabit Ethernet | Wi-Fi 6 (8m, 1 wall) | Wi-Fi 5 (8m, 1 wall) |
|---|---|---|---|
| Idle Ping | 4 ms | 12 ms | 18 ms |
| Loaded Ping (bufferbloat) | 6 ms | 47 ms | 89 ms |
| Jitter | 0.3 ms | 4 ms | 11 ms |
| Packet Loss (2.4 GHz environment) | 0% | 0.2% | 1–3% |
| Speed (download) | 940 Mbps | 620 Mbps | 380 Mbps |
Why Loaded Latency Is What Actually Matters
Notice the loaded ping column. Idle ping — the number you get from a basic ping test with no other traffic on the line — looks deceptively reasonable for Wi-Fi. But the moment you add real traffic (someone else streaming, a background app syncing, a game download running), wireless latency deteriorates sharply. This is called bufferbloat, and it is the real reason your game lags right after someone else in the house starts watching Netflix. Ethernet does not do this because the medium does not have to negotiate access on every frame. It just sends. Wi-Fi, by protocol design, has to wait its turn.
When Wi-Fi Is Actually Fine
I am not saying Wi-Fi is worthless. For the majority of use cases — casual browsing, streaming video, video calls on a quiet network — modern Wi-Fi is genuinely excellent. The 12ms Wi-Fi idle ping is functionally invisible to a human on a video call. Where the difference becomes real and measurable is:
- Competitive online gaming — where 40ms extra loaded latency is the difference between winning and losing a gunfight
- Real-time audio/video production — where jitter causes audio dropouts and sync issues
- Multiple simultaneous heavy users — where the shared wireless medium degrades for everyone at once
- Remote desktop and VPN work — where jitter makes the experience feel sluggish even at adequate speeds
The $8 Fix That Beats a $400 Router
If you fall into any of those categories above, the answer is not to buy a better router. It is to run a cable. A Cat6 Ethernet cable costs about eight dollars for a 15-meter length. It will deliver deterministic, jitter-free performance that no wireless technology can match by fundamental physics. Your 5 GHz signal has to share the same unlicensed spectrum with your neighbors' routers, their microwaves, baby monitors, and Bluetooth devices. Your Ethernet cable shares spectrum with nothing.
If running a cable is genuinely impossible — through finished walls, multiple floors — then a powerline Ethernet adapter or a wired mesh backhaul is the next best option. A Wi-Fi extender is the worst option; it adds another layer of wireless retransmission and typically makes latency worse, not better.
How to Measure Your Own Setup
Do not take my word for it. Run a speed test on your current connection, note your ping and jitter, then plug in directly and run it again. The numbers will tell you everything you need to know about whether your specific situation warrants the cable run. In most cases, they will make the decision for you.