The talk test, explained

Why it works

Speaking requires controlled exhalation. You take a breath and spend it gradually, shaping airflow into words.

Below your first ventilatory threshold (VT1), breathing is unhurried enough that you can borrow air for speech without disrupting it. Above VT1, ventilation climbs disproportionately as your body works to clear accumulating carbon dioxide. Breathing takes priority over speaking, and speech fragments around it.

That boundary — VT1 — is the physiological event that "zone 2" is trying to name. The talk test detects it directly through its consequences, which is why coaches have relied on it for decades without any equipment at all.

Ventilation also responds to that metabolic shift quickly, while heart rate takes minutes to settle at a new workload and drifts upward over long efforts even at constant pace. Speech often reflects the change first.

How to run one

1. Warm up properly — several minutes of easy movement. Testing on cold legs gives a misleading answer.
2. Settle at a steady effort and hold it long enough that your breathing stabilises. A minute or two is reasonable.
3. Speak continuously for about thirty seconds. Recite something you know by heart, or count from one to thirty in a normal speaking voice — not a whisper, not a shout.
4. Judge the result honestly. Could you comfortably keep going? Or were you breaking sentences to breathe?
5. Note your heart rate at the last effort where speech was still fluent. That is far closer to your real zone 2 ceiling than any formula's estimate.
6. Repeat at a slightly harder effort to find where fluency breaks down. The boundary between the two is your threshold.

What "breaking down" actually sounds like

The change is gradual, so it helps to know what you are listening for:

• Shorter phrases — you break sentences in places you normally would not.
• Breaths interrupting words — audible gaps mid-phrase rather than at punctuation.
• Rising pitch — vocal cords tighten under strain.
• Uneven rhythm — speech becomes rushed and clipped rather than steady.

Getting it wrong

Two failure modes account for most bad results.

Testing too briefly. A few words is not a talk test. It takes twenty to thirty continuous seconds for breathing to genuinely interfere — a short sentence can be forced out well above threshold.

Judging generously. "I could talk if I had to" is not the standard. The standard is comfortable, sustainable conversation. If you are consciously managing your breathing to get words out, you are above the line.

Wind, cold air, hills, and talking while running downhill will all skew a reading. Test on flat, steady ground where you can.

Automating it

The talk test's virtue is that it needs no technology — which makes an app that automates it a slightly odd proposition, and worth being straight about. The reason to automate is not that speaking is hard, but that judging your own speech is: the changes are subtle, you are the least objective observer of your own breathlessness, and the interesting question is the exact heart rate where fluency breaks down.

RunEasy runs a guided ramp on Apple Watch: a warm-up, then stages stepping the effort up, with a spoken question near the end of each. It measures phrase length, breath interruptions, pitch elevation and rhythm to locate where speech degrades, and reports the corresponding heart-rate band — alongside what the age formula would have predicted.

Used once, that gives you a measured number to train against. After that you do not need to talk to your watch at all.

Sources

• High North Running — "The talk test and aerobic threshold for runners"
• Should I Train — "Is zone 2 training overrated?" — on where zone precision matters and where it does not.
• Steve Magness — "Stop arguing about zone 2" — argues the classical talk test is the practical standard.