📚 Part of our Heart Health Guide — an evidence-based guide.
Blood pressure is the most consequential number in preventive medicine, and it is also the one most likely to be measured wrong. Not slightly wrong — wrong by enough to change whether you are told you have hypertension.
The errors are not exotic. They are things like where your arm was resting, whether you had spoken in the previous minute, and whether the cuff was the right size. Each carries a documented effect on the reading, and they add together.
Why the office reading is the weaker measurement
For most of medical history, blood pressure meant what a clinician measured in a clinic. That has changed. Major guidelines now recommend out-of-office measurement to confirm a diagnosis of hypertension, because home and ambulatory readings predict cardiovascular events better than office readings do.
Two mismatches explain why.
White coat hypertension — high in the clinic, normal at home — affects a substantial share of people with elevated office readings. Treating it as sustained hypertension means treating a number that does not exist most of the day.
Masked hypertension is the reverse and the more dangerous one: normal in the clinic, high the rest of the time. It carries cardiovascular risk comparable to sustained hypertension, and by definition an office visit will never find it. It is more common in people with diabetes, chronic kidney disease, or high-stress work.
A home monitor is the only practical way most people will detect either.
The errors, and what each one costs
These come from validation studies and are reported as approximate effects on systolic pressure. Individual variation is wide, but the direction is consistent.
Wrong cuff size — up to 20 mmHg. A recent randomised crossover trial found that using a regular cuff on an arm requiring a large one overestimated systolic pressure substantially, while an oversized cuff on a smaller arm underestimated it. This is the single largest correctable error, and it is the one most people never check. The bladder should encircle about 80% of the arm. Measure your mid-upper arm circumference and buy for it.
Unsupported back — roughly 6 to 10 mmHg. Sitting on an exam table with no back support raises the reading. Use a chair.
Legs crossed — roughly 2 to 8 mmHg.
Feet dangling — roughly 6 mmHg. Flat on the floor.
Arm hanging or unsupported — up to 10 mmHg. The cuff must be at heart level. A randomised trial published in 2024 found that resting the arm in the lap, or letting it hang at the side, produced markedly higher readings than supporting it on a desk — enough to misclassify people.
Cuff over clothing — roughly 5 to 50 mmHg, depending on thickness. Bare arm.
Talking — roughly 10 mmHg. Including answering a question from whoever is taking it.
Full bladder — roughly 10 mmHg.
Stack four of these and the total error exceeds the entire distance between normal and stage 2 hypertension (what the numbers mean).
The protocol
Thirty minutes before: no caffeine, no exercise, no smoking. Empty your bladder.
Sit for five minutes in a chair with back support, feet flat, legs uncrossed, in a quiet room. Do not use a phone. This rest period is the step most often skipped and it is not optional.
Place the cuff on a bare upper arm, lower edge about 2–3 cm above the elbow crease, snug enough that two fingers slide under. Rest the arm on a table so the cuff sits at heart level.
Stay silent during the measurement.
Take two or three readings, one minute apart, and average them. Discard the first if you have three — the first is typically the highest. A single reading is not a measurement.
Do this morning and evening for seven consecutive days before any appointment where the number will matter. Then average everything except day one. That seven-day average is the value guidelines are built around — not a single alarming reading.
Choosing a monitor
Buy an upper-arm device. Wrist and finger monitors are more sensitive to position error and are generally not recommended for diagnosis.
Buy a validated device. Most monitors on the market have never passed independent validation against a standard protocol. Independent listings — validatebp.org in the U.S., stridebp.org internationally — publish which models have. Check the list before buying; it costs nothing and the price difference is often zero.
Take your monitor to your next appointment and compare it against the office device on the same arm within a few minutes. That single check tells you whether to trust it.
Reading your own numbers honestly
Home thresholds are slightly lower than office ones, because the clinic effect is absent. Home hypertension is generally defined at 135/85 mmHg or above on average, against 140/90 in the office under most guidelines — and lower still under the ACC/AHA definition, which sets hypertension at 130/80.
Blood pressure varies through the day by 10 to 20 mmHg in healthy people, and that is normal physiology. One high reading is not a diagnosis and not an emergency. The average over a week is the signal; the individual readings are mostly noise around it.
Resist the temptation to re-measure repeatedly after a high number. Anxiety about the reading raises the reading.
And do not adjust medication on your own based on home numbers. Bring the log — that is what it is for.
When to see a doctor
Bring a seven-day log if your average is at or above 135/85. Ask specifically about masked hypertension if your office readings are normal but you have diabetes, kidney disease, or a strong family history — because that is the situation office measurement is worst at detecting.
Seek emergency care for a reading of 180/120 or higher accompanied by chest pain, breathlessness, vision change, weakness on one side, difficulty speaking, or severe headache. Without those symptoms, a very high reading still warrants prompt medical contact — but repeat it correctly first, because a badly taken measurement produces exactly these numbers.
Hypertension is largely symptomless, which is the entire reason measurement matters (the honest answer on symptoms).
The honest bottom line
A home monitor predicts your cardiovascular risk better than the reading taken at your appointment — but only if the technique is right, and most of the time it is not. Cuff size, arm support, and a genuine five minutes of sitting still account for the majority of the error, and together they can move a reading further than most first-line medications do. Get a validated upper-arm device, run seven days of paired mornings and evenings, average it, and bring that number in. It is worth more than anything measured in a hurry.
Frequently asked questions
What is the correct way to measure blood pressure at home?
Avoid caffeine, exercise and smoking for 30 minutes and empty your bladder. Sit five minutes in a chair with back support, feet flat, legs uncrossed. Put the cuff on a bare upper arm, about 2–3 cm above the elbow crease, resting on a table at heart level. Stay silent. Take two or three readings a minute apart and average them, morning and evening for seven days.
Why is my blood pressure higher at the doctor's office?
This is white coat hypertension, and it affects a substantial share of people with elevated office readings. The reverse also exists: masked hypertension, normal in the clinic and high the rest of the time, which carries risk comparable to sustained hypertension and cannot be detected in an office visit at all. It is more common with diabetes, kidney disease or high-stress work.
Are wrist blood pressure monitors accurate?
Upper-arm monitors are recommended for diagnosis. Wrist and finger devices are far more sensitive to position error, since small changes in height relative to the heart alter the reading substantially. If you already own one, compare it against a validated upper-arm device before relying on it.
What blood pressure is considered high at home?
Home thresholds are lower than office ones because the clinic effect is absent. Home hypertension is generally defined as an average of 135/85 mmHg or above, against 140/90 in the office under most guidelines — and the ACC/AHA definition sets hypertension at 130/80. What matters is the seven-day average, not any single reading.
How do I know if my home blood pressure monitor is accurate?
Check whether the model appears on an independent validation listing such as validatebp.org or stridebp.org — most monitors on the market have never passed independent validation, and validated models are often no more expensive. Then take your device to your next appointment and compare it against the office monitor on the same arm.