How to Test Your Internet Speed Accurately on WiFi and Ethernet

A broadband plan’s advertised speed is only the starting point. To test your internet speed accurately on WiFi and Ethernet, remove avoidable interference, choose a suitable test server, and measure the connection more than once. A single quick reading can reflect a busy household, a distant WiFi access point or momentary server load rather than your usual service.

WiFi and Ethernet measure different parts of your setup. Ethernet usually provides the cleanest view of the line delivered by your internet service provider, while WiFi adds signal strength, congestion, walls and radio interference. Testing both helps identify whether a slow result comes from broadband itself or your local wireless network.

Prepare for a reliable test

Start with a desktop or laptop connected by Ethernet where possible, then compare that result with WiFi. Close streaming apps, cloud photo sync, game downloads and large file transfers. Pause software updates, disconnect a VPN and stop any backups already running. Restarting the router and modem a few minutes beforehand can clear stale sessions, although frequent reboots are not a substitute for troubleshooting.

Australian providers such as Telstra, Optus and TPG may supply different routers and network equipment. Record your plan speed, connection type and router model. Test at a quiet time first, then repeat during the evening busy period, often around 7–10 pm in Australia. Turn off unused devices, especially if they are watching 4K video or uploading files. For a mesh system, begin near the main router with satellite nodes switched off.

Choose the right speed test tool

Ookla Speedtest is a practical choice for Australian users and is available in a browser and as a desktop application. Choose an Australian server, preferably one on your provider’s network when available, and use the same server for repeat measurements. A nearby server may reduce travel time, but a congested route can still produce inconsistent results.

Run three to five tests, pausing briefly between each, and compare the median download and upload figures rather than the highest number. Close other heavy programs and use the same device for every run. If possible, repeat on another day: consistency matters more than an impressive peak achieved once.

Test your WiFi connection

Place the device close to the router for the first test, ideally in the same room. Connect to the 5 GHz or 6 GHz band when available; these bands generally handle crowded apartment blocks better than 2.4 GHz. The 2.4 GHz band offers longer range but is more exposed to neighbouring networks. Check your router settings to confirm which band you are using.

Move farther away and test again from places where you normally use the connection, such as a bedroom or home office. Walls, floors and microwaves can weaken a signal, while neighbouring access points can reduce throughput. If the router reports signal strength or link speed, remember that a strong signal does not guarantee fast internet: the provider’s line and the router’s processing limits still apply.

Test your Ethernet connection

Connect the computer directly to the router or modem with a Cat5e or Cat6 Ethernet cable. Bypass a mesh node, powerline adapter and extra switch for the baseline test. On a gigabit connection, check that the network adapter and router port negotiate at 1,000 Mbps; a 100 Mbps link identifies a cable, port or adapter limitation. Restart the computer after changing cables if the negotiated rate remains incorrect.

Repeat the speed test with the wired device replacing the WiFi device. A strong wired result paired with much slower WiFi results points towards the wireless side, while similarly slow results suggest the broadband line, router or provider route. Compare like with like: use the same test server, testing period and number of runs for each connection method.

Interpret the results in context

Download speed shows how quickly data reaches you, upload speed shows how quickly it leaves, ping measures response time in milliseconds and jitter describes how much response time varies. For streaming and browsing, stable downloads matter; for gaming and video calls, latency and packet loss are often more noticeable than headline speed. A brief reading may look healthy while a file transfer stalls, so inspect sustained performance when a service is genuinely failing.

Access technologies differ across Australia. The NBN may use fibre-to-the-premises (FTTP), fibre-to-the-node (FTTN), fibre-to-the-curb (FTTC) or fixed wireless, with speeds shaped by the technology and time of day. Some apartments in Sydney and Melbourne still have Foxtel cable, while regional Western Australia may rely on fixed wireless or satellite services such as Starlink. These networks have different latency, capacity and peak-hour behaviour, so compare results with your plan’s expectations rather than a maximum figure advertised elsewhere.

Know what to do about a poor result

If the result is lower than expected, repeat the test after closing background activity, try another server and test again over several days. Send your provider the plan name, connection details, dates, times and screenshots. Ask them to check the line, speed profile and recent network faults. A result below the plan may still indicate a fault, but one isolated reading is rarely enough to establish one.

Keep records rather than relying on memory. The Australian Consumer Law provides guarantees that services are supplied with due care and skill and are reasonably fit for their disclosed purpose, while misleading speed claims can breach consumer law. These protections do not promise the fastest possible connection every day, but they matter when a service fails its stated purpose or a provider’s representation is inaccurate. If the provider will not investigate, a complaint to the Telecommunications Ombudsman may be appropriate for eligible consumers.