Not for individuals. It was published in 1970, was never derived from original research, and carries a standard deviation of roughly 10–12 beats per minute. Two people the same age can have genuine maximum heart rates more than 20 bpm apart. It describes a population average tolerably well and any particular person poorly.
The formula is usually credited to Dr William Haskell and Dr Samuel Fox around 1970. It was not the output of a dedicated study designed to predict maximum heart rate. They compiled data from roughly eleven existing published and unpublished sources, plotted it, and drew a line describing how maximum heart rate declines with age on average.
It was offered as a rough guideline. Haskell himself later expressed surprise at how literally it came to be taken. Its durability owes more to being memorable than to being accurate — it is easy arithmetic you can do in your head, which is exactly the property that made it spread into every treadmill console and fitness watch on the market.
The commonly cited standard deviation is about 10–12 bpm. In practice that means:
The error compounds when you build zones on it. Zone 2 is typically defined as 60–70% of maximum, so a 15 bpm error in the maximum becomes roughly a 10 bpm error in your zone 2 ceiling — which is easily the difference between a comfortable jog and being forced to walk.
| Method | What it improves | What it does not fix |
|---|---|---|
| 220 − age | Nothing; it is the baseline | ±10–12 bpm individual spread |
| Tanaka: 208 − (0.7 × age) | Fits modern data better, especially for older adults | Still a population regression; individual spread remains |
| Karvonen (heart-rate reserve) | Uses your resting heart rate, so it partly personalises the range | Still needs a maximum, which is still estimated |
| Measured threshold (talk test, lactate, gas exchange) | Measures you, no population average involved | Requires a test; thresholds drift as fitness changes |
The pattern is worth naming: every formula-based method is trying to guess a number that can simply be measured. Better formulas narrow the error. Measurement removes the guess.
It is not useless. It works acceptably for:
What it is not suited to is what most people use it for: prescribing a precise heart-rate ceiling and then obeying it during every run.
Measure a threshold rather than estimating a maximum. The talk test is the accessible option and needs no equipment: the effort at which continuous speech starts to break down corresponds closely to the first ventilatory threshold, which is the boundary zone 2 is trying to identify. Lab testing (lactate or gas exchange) is more precise if you have access to it.
Compare all four approaches → · Read the full guide →
A note on training. This is general training information, not medical advice. Maximum-heart-rate field tests are strenuous by design; if you are new to exercise or have a heart condition, talk to your doctor first.