How to test for bufferbloat in 5 minutes (step by step)

What a bufferbloat test actually measures

A normal speed test measures how much data your line carries. A bufferbloat test measures something else: how long your packets wait in line when the line is busy. A router with too much queued data holds packets for a long time once a download fills the link, so ping jumps even though the speed looks fine. That is why a test that only reports Mbps can miss the problem a game is having.

The test records two numbers. Idle ping is the delay when nothing else is running. Ping under load is the delay while the test pushes download and upload traffic. The gap between them is the number that matters for games and video calls.

Set up the test in two minutes

1. Connect one computer to your router with an Ethernet cable. Wi-Fi adds its own delay and jitter, so it hides the result you're trying to measure.
2. Close every download, cloud sync, game update, video call and VPN on every device in the house. Pause streams too. A test on a busy network tells you about the people using it, not your router.
3. Make sure the test computer isn't busy. If its fans spin up during the test, close other programs first.
4. Run the test at the time you actually play. Evenings in the US are when most homes share the line, so a 9 pm result is more useful than a 10 am one.

Run the test and read the two numbers

Use the Net Fix speed and lag test, which grades ping under load from A+ to F, or any well-known bufferbloat test. Run it twice, a few minutes apart, so one busy moment doesn't decide the result.

What to do with a bad result

If the rise is 30 ms or more, the usual fix is Smart Queue (SQM), which shapes traffic just below your line speed so the router never builds a long queue. Some routers have it built in, and the router guides show where the setting is. The bufferbloat fix guide walks through the settings step by step.

If the result is still bad at the same time every evening, test again at a few different hours on a cable. Your last results stay in this browser, so you can see whether the problem is your home or the line at peak hours, and show the numbers to your provider.

Five mistakes that give a false all-clear

Most people who run a bufferbloat test and see a good grade are testing the wrong thing. These five mistakes are the usual cause, and each one hides the queue that makes games stutter.

1. Testing on Wi-Fi from the router's own room. Strong signal hides the problem, because the packets never queue on the air. Test from the room where you play, and wire the test computer if you can.
2. Leaving a background download running. A cloud backup or console update fills the line for you, and the queue is measured while it's already slow. Pause everything and check the task manager or network tab.
3. Testing only once in the morning. A line can be clear at 10 am and jammed at 9 pm. One result at the wrong hour proves nothing about the evening.
4. Trusting a grade without the two numbers. An A on a slow plan and a C on a fast one can both be the same problem. The rise in ping under load is what tells you if games will stutter.
5. Confusing bufferbloat with a bad server. A game server that is far away has a high idle ping, and that is not bufferbloat. Bufferbloat is the rise between idle and loaded, not the height of the idle number.

Check your router's queue settings

Your router decides how long packets wait when the line is full, so its settings matter as much as the test. Many US home routers ship with no queue management turned on, which is why the rise under load is often the first thing to appear after a new router is set up. A bufferbloat test before and after you change a setting is the only reliable way to know the change helped.

Look for Smart Queue, SQM, fq_codel, CAKE or QoS in the advanced settings of the router's app or web page. Set the download and upload limits a little below what you measured on a clean test, not the plan's advertised speed. Retest after each change. If the rise falls under 30 ms and stays there at 9 pm, the setting is working.

Read the result against a real game session

A grade is easier to act on when you tie it to what you play. Think about a match of a fast shooter. Your ping has to stay steady for 20 or 30 minutes, and a spike of 150 ms during a fight is what makes you lose the round, even if your average is fine. A bufferbloat rise of 80 ms means that spike can happen whenever someone else in the house starts a download, which is exactly the pattern people describe as random lag.

Run the test while you play, too. Start a match, open a second window that runs the test, and compare the ping under load with what your game's network display shows. If the two line up, the queue is your problem. If the game ping is steady while the test shows a big rise, the cause is probably the game server or matchmaking, and the game lag guide covers that side.

For video calls the target is different. Calls use less bandwidth, so they are mostly hurt by jitter and by the rise on the upload side. Run the test once with an upload running and once without. An upload-side rise over 30 ms is the one that makes a call freeze or cut out when someone in the house backs up photos.

Keep a note of each result with the time, the device, and what was running. Three notes over a week are usually enough to show a pattern, and that pattern is what a technician or your provider will need if the router settings don't fix it.

How often to test

Test once after you set the router up, once more after any firmware update or new device, and during the evening when you notice lag. Results change with the time of day, so one good test at noon doesn't rule out a problem at 9 pm.