How to Measure Your Own Blood Pressure at Home

In over twenty years of running a hypertension service, I have come to believe that the single most useful skill a patient with raised blood pressure can acquire is not a piece of knowledge about salt, alcohol or medication — it is the ability to measure their own blood pressure accurately at home, and to summarise the result in a form their GP or specialist can actually act on. Done badly, home measurement produces wildly variable numbers that confuse rather than clarify. Done well, it gives both patient and doctor an objective signal that no single clinic reading can match.

Raised blood pressure is still sometimes called the silent killer. Roughly a third of patients with hypertension do not know they have it, in a further third it is poorly controlled, and only the remaining third receive adequate treatment^[1]. Untreated mild hypertension in midlife is now the leading modifiable risk factor for heart failure hospital admissions in the elderly, and contributes to stroke, vascular dementia, atrial fibrillation and chronic kidney disease^[2].

The gold standard for diagnosis recommended by NICE remains 24-hour ambulatory blood pressure monitoring (ABPM)^[3]. For day-to-day management of known hypertension, structured home blood pressure monitoring (HBPM) using a validated arm-cuff device and a disciplined methodology is cost-effective and equally informative^[4]. The current NICE threshold for hypertension on home readings is 135/85 mmHg — slightly lower than the 140/90 mmHg used in clinic, because home readings tend to run a few mmHg lower than those taken in a medical environment.

There are hundreds of devices on the market, and the only sensible approach is to use one that has been independently validated, ideally by the British and Irish Hypertension Society or one of its sister organisations^[5]. Validated lists are freely available online. Costs typically range between £35 and £100, and the more expensive devices in this range are not appreciably more accurate than the cheaper ones — what matters is independent validation, not price.

Two non-negotiable points

Where possible, the practice nurse at your GP surgery should check your device against a manual or clinical-grade reading to confirm its accuracy.

For patients whose anxiety affects the reading

A small but real subgroup of patients — perhaps one in twenty — find that the act of taking their own blood pressure pushes the reading up, sometimes substantially. For this group there are two alternatives. The first is 24-hour ABPM, available through most cardiology services and the NICE-recommended gold standard for diagnosis. The second is the Hilo wristband (originally marketed as Aktiia), a relatively new optical device validated against arm-cuff measurement, which takes blood pressure readings passively without the patient consciously triggering each one^[8]. The Hilo has proved popular with our patients and currently costs about £200 for the first year; however, the company has recently introduced a substantial annual subscription fee (currently around £100 per annum thereafter), which makes it an expensive, albeit convenient, long-term option. For the great majority of patients, a validated arm-cuff device used with the methodology described below remains the best balance of accuracy, sustainability and cost.

Any individual blood pressure reading carries a measurement error of approximately ±5–8 mmHg. Layered on top of that is true physiological variation throughout the day, between days, and in response to whatever the patient happens to be feeling or doing in the minutes before the cuff inflates. In practice this means a one-off reading of, say, 148/92 mmHg might genuinely reflect an underlying average anywhere between roughly 138/86 and 158/98 — a range wide enough to either reassure or alarm, depending on which way the dice fall on the day.

The way to extract a meaningful number from a noisy signal is averaging — not picking the lowest of three, not relying on a single morning value, but a disciplined weekly mean.

If you take repeated measurements at the same time each day for between two and seven days and average them, the random errors largely cancel out — a statistical phenomenon known as regression to the mean^[9] — and what is left is something very close to your true underlying blood pressure.

Blood pressure is not constant through the 24-hour cycle. In approximately 90% of the patients I see, the highest reading of the day occurs in the morning, as part of the normal cortisol-and-aldosterone wake-up surge^[10]. In the remaining 10% or so, the pattern is inverted, and the highest readings occur in the evening. This matters because if a patient with the more common morning-high pattern measures only in the evening, they will derive a falsely reassuring picture and may be undertreated.

A further reason to favour the morning is medication timing. In patients already on antihypertensive therapy, the morning reading taken before the first dose reflects the lowest point of drug cover — exactly the reading that matters for deciding whether the regime is adequate.

Standardise the conditions. The setting I recommend is sitting at the kitchen table, feet flat on the floor, arm supported at heart height, cuff over bare skin, while the kettle boils. Keep the device at the same height each time — taking readings with the device on your lap one day and on the table the next will introduce spurious variation.

For reasons that are not fully understood — and which probably reflect a mixture of cuff adjustment, postural settling and a small startle response — the first reading is almost always different from the second and third, and is usually higher. The second and third readings are typically much closer to each other and more reflective of the true value, which is why my method discards the first and averages the second and third.

Common mistakes to avoid

Doing the arithmetic by hand each week — discarding firsts, averaging seconds and thirds, then averaging across days — is the part where most patients give up. We have therefore built a free web-based calculator, MyBP, that does it for you. Enter your readings each day; the tool produces your daily averages and your weekly home blood pressure automatically, and generates a clean summary you can copy or email straight to your GP or cardiologist.

The NICE and British Hypertension Society thresholds for hypertension on home readings are 135/85 mmHg. Clinic readings use 140/90 mmHg because they tend to run a few mmHg higher than home readings, largely due to the well-documented white-coat effect^[11].

If your weekly home average is below 135/85 mmHg and you feel well, repeat the exercise in 6–12 months, or sooner if you are running an N-of-1 lifestyle experiment — a two-week salt reduction trial, a structured aerobic exercise programme, or a weight-loss intervention.

If your weekly average is between 135/85 and 150/95 mmHg and you are not yet on medication, this is the sweet spot in which non-pharmacological measures — weight reduction, salt reduction, alcohol moderation, regular aerobic exercise, and in selected patients beetroot juice or other dietary nitrate-rich strategies^[13] — are most likely to be sufficient to bring you back below threshold. Repeat the home BP exercise after four to eight weeks of any lifestyle change to see whether it has actually worked.

If your weekly average is above 150/95 mmHg, or if you are already on medication and the number has not come down to target, send the MyBP summary to your GP or cardiologist. Modern antihypertensive medications — ACE inhibitors, angiotensin receptor blockers, calcium channel blockers, indapamide and spironolactone — are extremely well tolerated, exceptionally well evidenced and astonishingly cheap. Reluctance to start them is understandable but, in my experience, almost always misplaced once the patient has seen a full week of objective home readings demonstrating that the problem is real.