A home blood-pressure reading can be wrong by more than the difference that decides whether you are treated — and every one of the reasons is something you can fix this evening.

Your reading, and what your routine did to it

Put in the last reading you wrote down, then describe how you actually took it — honestly, not how the leaflet says. Each answer contributes a known, signed shift in millimetres of mercury, and the tool adds them up and shows the corrected estimate with the uncertainty it deserves. Nothing is saved, nothing leaves the page, and this cannot diagnose anything or tell you what to take.

Why a number that feels objective is not

• Blood pressure is not a property, it is a moment. It changes with every breath, rises on waking, falls in sleep, and swings by twenty or thirty millimetres across an ordinary day. A single reading samples one moment out of thousands, which is why no guideline anywhere diagnoses hypertension from one number.
• The measurement adds its own error on top of that. Cuff size, arm position, posture, talking, a full bladder and the time since you last moved each shift the result in a known direction by a known rough amount. They are not mysterious: they are physics and physiology, and they add up.
• The errors are the same size as the decisions. The gap between a home average that gets treated and one that does not is around five to ten millimetres. A single mistake in technique is that big on its own, so a careless routine does not blur the answer — it can change it.
• They do not all point upwards. Taking readings until a low one appears, using a wrist device, or measuring only when you feel calm can hide real hypertension. That is the more dangerous direction, because nothing about a falsely low number ever prompts anybody to look again.
• The first reading of a session is nearly always the highest. Attaching a cuff to a person raises their pressure; the second and third readings settle. That is why the standard advice is to take two or three a minute apart and discard the first, and why a single reading is systematically high.
• Averages are what the thresholds refer to. The numbers in every guideline are averages of a week of readings, not a personal best or a worst. Comparing one reading to a threshold built from averages is comparing two different kinds of thing.

The cuff, the arm, and the two arms

• A cuff that is too small reads high, and it is the commonest equipment error there is. The bladder inside the cuff has to wrap roughly 80% of the way around the upper arm. If the arm is large and the cuff is the standard one that came in the box, the reading can be eight to fifteen millimetres too high — enough to put somebody on tablets they do not need.
• An upper-arm cuff, and a validated monitor. Wrist and finger devices are convenient and much less reliable, because the wrist has to be held at exactly heart height and the arteries there are small. Manufacturers' validation lists exist; a device that is not on one is a guess with a digital display.
• Bare arm, cuff at heart height, arm resting on something. An arm held up, or left hanging, or resting in your lap below heart level, changes the reading by up to eight or ten millimetres, because you are partly measuring the height of a column of blood rather than the pressure in it.
• Check both arms once, then always use the same one. A persistent difference of more than about ten millimetres between arms is itself worth investigating, and once you know which arm reads higher, that is the arm to use from then on. Almost nobody is told this, and it is a single one-off measurement.
• Feet flat, back supported, legs uncrossed. Crossing your legs adds a few millimetres; sitting on the edge of a bed with nothing behind you adds a few more. These are small individually and they are in the same direction, which is how a routine drifts into being wrong by ten.
• The tube should not be over your clothing, and the cuff should not be over a jumper. If you cannot get to a bare upper arm, that is worth mentioning when you report the numbers, not silently ignoring.

The routine that produces a number worth reporting

1. Empty your bladder first. A full bladder adds around ten millimetres, and it is the single easiest correction on this page.
2. Sit quietly for five minutes, with nothing to read and nobody to talk to. Not one minute. The fall over those five minutes is real and it is the difference between a resting pressure and an arriving one.
3. No coffee, tea, or cigarette in the previous half hour, and no measurement straight after exercise or a stressful phone call.
4. Take two, ideally three readings a minute apart, and write down the average of the last two. Discarding the first is not cheating: it is the recommendation, because the first is inflated by the cuff going on.
5. Same arm, same time of day, twice a day for seven days when it matters. Morning before medicines and food, evening before bed. Discard the whole of the first day and average the rest — that is the number a clinician can actually use.
6. Write down every reading, not the ones you like. A monitor's memory is better than a notebook precisely because it cannot be edited, and a selective record is worse than no record because it looks like evidence.

Why the clinic number and your number disagree, and which one counts

• Home and 24-hour readings predict strokes and heart attacks better than clinic readings do. That is why guidelines now ask for home or ambulatory monitoring before starting lifelong treatment, and why your own careful average is not a second-class measurement.
• White-coat hypertension is high in the clinic and normal at home, and it is common. Treating it as ordinary hypertension means medicating somebody on the strength of the setting they were measured in.
• Masked hypertension is the opposite and it is the dangerous one: normal in the clinic, high in ordinary life. It carries a risk close to sustained hypertension and it is invisible unless somebody measures outside the clinic. A comfortable routine that produces reassuring numbers is exactly how it stays hidden.
• A 24-hour monitor settles the argument. It measures while you work and while you sleep, and the overnight readings carry information nothing else does — a pressure that does not fall at night is a finding in its own right. If the clinic and your home readings disagree, that is the test to ask about by name.
• Home thresholds are lower than clinic thresholds, on purpose. Roughly 135/85 at home corresponds to about 140/90 in a clinic, because the clinic itself adds a few millimetres. Comparing a home average to a clinic threshold makes it look better than it is.

What the two numbers mean, and what they are not

• The top number does most of the predicting after about fifty. Systolic pressure is the one that tracks stroke and heart risk in later life; a raised top number with a normal bottom number is a real finding and not a technicality.
• High blood pressure has no symptoms until it has done damage. Headaches, nosebleeds and flushing are not reliable signs of it, and feeling well is not evidence. This is why it is measured rather than felt.
• The risk is continuous, not a cliff at any threshold. There is no number above which you are ill and below which you are safe; the thresholds exist to organise decisions. That cuts both ways: 138 is not a catastrophe, and 132 is not a licence.
• The treatment decision is about total risk, not the number alone. Age, diabetes, kidney function, cholesterol and smoking all change what a given pressure means, which is why no page and no app can tell you whether you should be treated.
• Most of the benefit comes from the first medicine and from consistency. The biggest gains are in taking something reliably rather than in finding the perfect combination, and in salt, weight, alcohol and movement — which change the number by amounts comparable to a tablet.
• A single low reading is not a reason to stop anything either. Pressure falls in the summer, during illness and after weight loss; a dose that has become too strong is a conversation and a plan, not a decision made alone on one evening's number.

When a number is not a technique problem

• 180/110 or above, repeated after five minutes of rest, needs medical advice the same day, whatever the routine was. No correction on this page applies to a number that high, and it is not something to average away over a week.
• The same day, urgently, if a very high reading comes with chest pain, breathlessness, weakness or numbness on one side, difficulty speaking, sudden severe headache, visual loss, or confusion. That is a stroke or heart pathway, not a blood-pressure diary.
• In pregnancy, the thresholds are different and lower, and a rise with headache, visual disturbance, upper abdominal pain or sudden swelling is urgent regardless of the exact figure.
• Symptoms when standing up matter as much as high numbers. Dizziness, greying vision or near-fainting on standing suggests the pressure is dropping too far — often from treatment — and is a common, treatable cause of falls. Measure sitting and again after a minute standing, and report both.
• A pulse that the monitor flags as irregular is a separate finding. Many monitors show an irregular-heartbeat symbol; it is worth mentioning, because atrial fibrillation both makes readings less reliable and is treatable in its own right.

The appointment, in six questions

The second question is the one that changes appointments. A week of properly taken readings converts a conversation about one frightening number into a decision based on evidence you brought.

What people believe, and what is so

The drill: 16 decisions

Sixteen ordinary moments — a frightening number after climbing the stairs, a cuff that came in the box, a relative who takes it five times and keeps the lowest, a dose skipped because the reading looked good. Most have an instinctive answer that sounds sensible and is wrong. Pick your move; every answer explains why.

The card

Print it, tape it next to the monitor, and fill in a week.

🖨️ Print the card