RealMbps·guide

Guide · updated May 2026 · 7 min read

Fiber speed test: how to measure correctly

A poorly executed speed test is worthless. Here's how to get a trustworthy number, what it means for your contract, and what to do if your connection isn't keeping its promises.

In this guide
  1. FTTH, FTTC, FWA, ADSL: what line you actually have
  2. How to test properly: a reliable procedure
  3. Mistakes that ruin a test (5 classics)
  4. What to expect for each technology
  5. What to do if you're below the guaranteed speed
  6. FAQ

FTTH, FTTC, FWA, ADSL: what line you actually have

Before measuring, you need to know what you're measuring. The acronyms in your contract drastically change expectations:

How to find out your technology? On your ISP's website, in the "line status" or "technical details" section. Or look at your modem: SC/APC fiber port = FTTH, RJ-11 phone port = FTTC/ADSL.

How to test properly: a reliable procedure

A measurement that has any value in case of dispute must be repeatable and run under controlled conditions. The standard procedure:

  1. Connect your computer over Ethernet directly to the modem (not the router, not a switch). Wi-Fi adds variables that make tests unreliable.
  2. Close everything: browser tabs, Dropbox, OneDrive, iCloud, Steam, automatic updates. Even a single torrent client in the background can skew the result.
  3. Power off other devices in the home or disconnect them from Wi-Fi. Smart TVs, speakers, and cameras silently consume bandwidth.
  4. Run the test 3 times a few minutes apart. If results vary by 20%+ between runs, you have a stability problem.
  5. Repeat across different time slots: morning (8-10am), afternoon (2-4pm), evening (9-11pm). Evening is the peak hour and the critical one for complaints.

On our tester RealMbps the test runs against Cloudflare with a server picked automatically based on your geolocation: no tricks, no "friendly servers" inside your ISP's network.

Mistakes that ruin a test (5 classics)

1. Testing over Wi-Fi on a 300+ Mbps line

On a 1 Gbps FTTH line, your ISP's router rarely exceeds 400 real Mbps over Wi-Fi. Stop there and you're measuring the router, not the connection. Always test over Ethernet first.

2. VPN active

NordVPN, ExpressVPN, ProtonVPN, even a corporate VPN: they all add an intermediate hop that reduces speed and increases ping. Disable during the test.

3. Slow device

An old PC with built-in 802.11n Wi-Fi physically can't go above 150 real Mbps, even on a 1 Gbps line. Same with entry-level smartphones. Use adequate hardware.

4. Outdated browser

Old Chrome/Firefox/Safari versions have limited HTTP/2 and handle multiple connections poorly. Update before testing.

5. Phone on cellular thinking it's on Wi-Fi

Happens more than you'd think: phone connected to home Wi-Fi but cellular kicks in for default routing. Verify in settings that Wi-Fi is actually being used.

What to expect for each technology

Real numbers you should see on a healthy line, over Ethernet, off-peak:

If you're consistently below 60% of expected, there's a problem. Common causes: distance from the cabinet (FTTC), obsolete router, area congestion, line fault. Don't rule out internal wiring either: a corroded telephone splitter can halve an FTTC line.

What to do if you're below the guaranteed speed

The minimum guaranteed speed is in your contract. If your measurements are consistently below, you have two paths:

  1. Formal complaint to the ISP: contact customer service requesting a written complaint. Most ISPs are required to respond within 30-45 days.
  2. Official measurement tool: many countries have a state-sanctioned tool (Italy: Ne.Me.Sys; UK: Ofcom's monitor; Germany: Breitbandmessung). Only their reports have legal standing.
  3. Regulator mediation: if the ISP doesn't resolve, escalate to the telecom regulator. Typically leads to penalty-free contract termination and proportional refunds.

Measure your line now

Free test, no signup, no tracking. If you don't have a baseline, start here.

Start the test →

FAQ

Why does the test show fewer Mbps than my contract promises?

ISPs advertise peak speeds in ideal conditions, measured at their first network node. The test measures what reaches your device after router, Wi-Fi, and protocol overhead. Differences of 10-30% are normal; beyond that, investigate.

Should I test over Wi-Fi or Ethernet?

Always test over Ethernet first to measure the line. Then over Wi-Fi to measure your home network. The gap tells you whether the issue is the connection or the router/Wi-Fi.

What is the minimum guaranteed speed?

In many countries, fiber contracts must declare a minimum guaranteed speed. If your tests are consistently below it, you typically have the right to terminate without penalty and receive a partial refund.

How many times should I test?

At least 3 measurements on different days and time slots. A single measurement has zero statistical value.