Home readings predict outcomes better than clinic ones — if taken properly. Cuff size, the seven-day protocol, and common errors.
A single reading taken in a clinic, after you have rushed to get there and sat in a waiting room, is among the least reliable numbers in medicine. Readings taken at home predict strokes and heart attacks better than clinic readings do — which is why guidelines increasingly rely on them.
That only holds if they are taken correctly. Most are not, and the errors are large enough to change whether someone is diagnosed with high blood pressure at all.
An upper-arm monitor. Not a wrist monitor, and not a smartwatch.
Wrist devices are extremely sensitive to the position of the wrist relative to the heart, and give inconsistent results. Watches and rings that claim to measure blood pressure are not accurate enough for clinical decisions, whatever the marketing says.
Check that the model is clinically validated. The British and Irish Hypertension Society publishes a list of validated monitors on its website — checking a model against it before buying takes two minutes and rules out a good deal of what is sold online. A validated monitor costs relatively little; an unvalidated one is worse than none, because it produces numbers people act on.
A cuff that is too small gives falsely high readings — sometimes by a substantial margin. Enough to have someone started on medication they did not need.
Measure around the middle of your upper arm with a tape measure, and check that figure against the range printed on the cuff. Many monitors are sold with a standard cuff that does not fit larger arms, and a large cuff usually has to be bought separately. If your arm circumference is near the top of the printed range, size up.
Take two or three readings, one minute apart. Discard the first, and average the rest. The first reading is almost always the highest, and treating it as the answer is a reliable way to worry yourself unnecessarily.
This is what is actually used to make decisions, and it is worth doing properly if a diagnosis is being considered.
Write them down, or use the monitor's memory. Bring the whole set, not a summary — the spread is informative in itself.
Home thresholds are lower than clinic thresholds, because the clinic setting itself raises readings. This surprises people who compare the two.
The equivalent clinic figure is 140/90. So a home reading of 138/88 is not "fine because it is under 140" — it is above the home threshold.
White coat effect — high in clinic, normal at home — is common and is exactly why home monitoring exists.
Masked hypertension is the reverse: normal in clinic, high at home. It is less well known, carries real risk, and is only ever found by measuring at home.
Seek urgent medical help for a reading of 180/120 or above together with any of: chest pain, breathlessness, a severe headache, visual disturbance, weakness or numbness on one side, difficulty speaking, or confusion.
A very high reading with no symptoms at all still needs assessment promptly, but is not usually an emergency — sit quietly and repeat it before acting.
Blood pressure is one component of cardiovascular risk rather than the whole of it. Alongside it, it is worth knowing your cholesterol, HbA1c and kidney function, and having a urine test for protein.
And if your blood pressure is high and difficult to control, obstructive sleep apnoea is a common and treatable underlying cause that is regularly missed — particularly if you snore and wake unrefreshed.

Clinically reviewed by Dr Mohammad Zubair Khan, GMC 7563469
Last reviewed
August 29, 2026
Ipsum dolor sit amet consectetur commodo aliquam augue duis aliquet ipsum donec tempus ac interdum enim.