How to Measure Blood Pressure at Home: A Step-by-Step Guide

• Pick an upper-arm cuff monitor. Wrist and finger devices are far more sensitive to position and generally less accurate.
• Check it's clinically validated. Look for validation registries such as validateBP.org (US), the STRIDE BP list, or a note that the device passed an ISO/AAMI or ESH protocol. Big brands like Omron, Beurer, Withings, and Braun have many validated models.
• Get the right cuff size. Measure your upper-arm circumference; a too-small cuff reads high, a too-large one reads low. Most monitors list the cuff range on the box.
• Bring it to one appointment. Asking your practice to check your device against theirs once is a quick calibration sanity check.

In the 30 minutes before a reading: no caffeine, no smoking, no exercise, and no heavy meal. Empty your bladder — a full one can add up to 10 mmHg. Then sit quietly for five minutes. Not scrolling-quietly: actually still, feet on the floor, not talking.

Measure at the same times each day, ideally morning (before medication and breakfast) and evening. Blood pressure follows a daily rhythm, so consistent timing is what makes readings comparable.

1. Sit correctly: back supported against the chair, both feet flat on the floor, legs uncrossed.
2. Bare arm: put the cuff on skin, not over a sleeve; a rolled-up tight sleeve is worse than a thin shirt — take the arm out instead.
3. Cuff position: 2–3 cm above the elbow crease, snug enough for two fingertips to slide under, with the tube running down the middle of the arm.
4. Arm at heart level: rest your forearm on a table so the cuff is level with your heart. An arm hanging at your side reads roughly 10 mmHg too high.
5. Stay still and silent during the measurement — talking adds up to 10 mmHg.
6. Take two readings, one minute apart, and record both. If they differ by more than 5–10 mmHg, take a third.
7. Record everything immediately — systolic, diastolic, pulse, date, time, and anything unusual. Or skip the transcription entirely and let an app read the monitor's display for you.

For a diagnosis-grade picture — before an appointment, after a medication change, or when readings first look elevated — international guidelines, including the UK's NICE guidance (which the NHS follows), the AHA in the US, and Hypertension Canada, recommend this schedule:

• Measure morning and evening for 7 consecutive days
• Each session: two readings, one minute apart
• Discard day 1 (first-day readings run high while the routine is new)
• Average everything else — that average is your number

A home average of 135/85 mmHg roughly corresponds to a clinic reading of 140/90. Bring the full log, not just the average — the spread and the morning/evening difference are informative too. Then compare your average against the blood pressure chart.

Also: don't measure only when you feel stressed or unwell. That samples your worst moments and produces a scary, biased picture — the schedule, not your mood, should decide when you measure.

Measure both arms once, on separate occasions if you can, then use the arm with the higher reading from then on and stay consistent. A difference of a few mmHg between arms is normal. A consistent difference of more than 10 mmHg systolic is worth telling your doctor about — it can point to narrowing in one of the arteries supplying the arm, and larger differences have been associated with higher cardiovascular risk. It is a conversation, not an emergency.

Cuff size matters at least as much as arm choice. Too small and the cuff has to squeeze harder to close the artery, so it reads high; too large and it reads low. Wrap a tape measure around your bare upper arm at the midpoint between shoulder and elbow, then check the number against your monitor's cuff range:

How tight? Snug enough that you can slide two fingertips under the edge, and no tighter. The cuff sits 2–3 cm above the elbow crease on bare skin, with the tube running down the middle of the inner arm. A cuff pushed over a rolled-up sleeve is the single most unpredictable error on the list — take the arm out of the sleeve instead.

An upper-arm cuff is the first choice, and guidelines say so plainly. Wrist monitors are not junk, but they are far less forgiving: the wrist arteries are smaller and sit closer to the surface, and the reading changes with the angle of your hand.

The dominant error is height. Blood pressure changes by roughly 0.8 mmHg for every centimetre a limb sits above or below heart level, so a wrist resting in your lap can read 10–15 mmHg too high, and a wrist held up near your shoulder too low. Some models refuse to measure until their built-in position sensor is satisfied — a feature worth paying for if you buy one.

A wrist monitor is a reasonable choice when an upper-arm cuff genuinely will not work: an arm circumference beyond the largest available cuff, a lymphoedema or dialysis fistula that rules the arm out, or a conical upper arm the cuff cannot grip. In that case: choose a clinically validated model, rest your elbow on a table and bring the wrist to the middle of your chest, keep your palm turned toward you, and check the device once against your practice's monitor. Our guide to wrist blood pressure monitors goes further into which models and situations work.

If your pulse picks up as you wrap the cuff, you are not imagining it. The anticipation of a measurement raises blood pressure in a lot of people — the same reflex behind white coat hypertension, just relocated to your own kitchen. Left unmanaged it inflates every reading you take, and then the high number makes the next measurement more stressful still.

What actually helps:

• Sit for the full five minutes. Most of the anticipatory rise fades in that window. Set a timer so you are not counting.
• Breathe slowly for a minute beforehand — a longer out-breath than in-breath, roughly six breaths a minute. This is the one relaxation technique with a measurable, if modest, effect on the reading.
• Do not watch the display. Turn the monitor away, or look out of the window. Watching the number climb during inflation is its own small stressor.
• Take two readings and use the average, discarding nothing. The first reading is usually the highest for exactly this reason.
• Measure on a schedule, not on a feeling. Checking because you feel odd samples your worst moments and produces a frightening, biased picture.
• Do not re-measure repeatedly chasing a lower number. If a reading is high, note it, and let the week's average answer the question.

If measuring at home makes you so anxious that the readings are useless, say so to your doctor — 24-hour ambulatory monitoring exists partly for this, and it takes the decision out of your hands entirely.

Nothing on the consumer market measures blood pressure without compressing an artery. Watches, rings and phone-camera apps that claim a cuffless reading are estimating from pulse timing or from colour changes in the skin, and no such device has been validated to the standard a clinic uses. Treat any number they produce as entertainment, not data — our guide to cuffless blood pressure covers what the research actually shows.

There is a real use for your phone's camera here, and it is a mundane one: instead of measuring your blood pressure, it reads the numbers your cuff already produced. Point Blood Pressure Journal at the monitor's display and it captures systolic, diastolic and pulse with a timestamp, so nothing depends on remembering a reading long enough to write it down. That is the difference worth knowing about — see measuring blood pressure with a phone camera.