Everything you need to know about blood pressure

Blood pressure is the force of blood against your artery walls, shown as two numbers. In the US, under 120/80 counts as normal. A week of home measurements tells you far more than one visit, and everyday habits can move it by several points.

If you have chest pain, shortness of breath, back pain, weakness, numbness, vision change or trouble speaking, call emergency services, whatever the number.

The short version

• The two numbers. Systolic (the top number) is the pressure while your heart beats. Diastolic (the bottom number) is the pressure between beats. Both are in millimeters of mercury (mmHg).
• The categories. In the US, normal is under 120 and under 80. Stage 1 starts at 130 or 80, and stage 2 at 140 or 90.
• How to measure. Take two measurements morning and evening for seven days, with your back and arm supported and the cuff on bare skin, then average them all. The full method is on how to measure at home.
• What moves it. The things that lower it are the DASH eating pattern, less sodium, exercise, weight loss if you carry extra weight, and less alcohol if you drink more than two a day. Each gives a few points, and some give up to about 11 mmHg. The ranked table is on how to lower your blood pressure.
• What lowering it does. Across 613,815 people in trials, each 10 mmHg drop in systolic pressure cut major cardiovascular events by about 20 percent.
• When to act. If your home week averages 130/80 or higher, book a visit. At 180 or 120 and up with no symptoms, rest five minutes and measure again. If it stays that high, call your doctor the same day. The emergency page has the decision table.

You can check your own numbers with the checker.

How we came to measure it

The first published measurement was done on a horse. In 1733 the English clergyman Stephen Hales described opening an artery in a mare’s leg and connecting it to a glass tube nine feet long. The blood rose 8 feet 3 inches. The method had no use in a clinic, and none of his animals survived, but it showed that arteries hold blood under a pressure you can measure.

The cuff doctors use today was invented in two steps. In 1896 the Italian physician Scipione Riva-Rocci introduced an inflatable arm cuff attached to a mercury column. It closed off the artery from outside, so no one had to cut anything, but it only gave the systolic number. In 1905 the Russian surgeon Nikolai Korotkoff reported that a stethoscope over the artery picks up sounds as the cuff deflates. Those sounds start and stop at two pressures, which gave doctors both numbers. The unit is still mmHg because those early cuffs used a mercury column.

For decades after that, doctors could measure the number but didn’t know what it meant, and moderately high pressure was often called benign.

How it works

Pressure depends on how much blood the heart pushes out and how much the arteries resist it. Push more blood or narrow the vessels and the number goes up.

• Systolic is the peak, as each heartbeat sends blood into the arteries.
• Diastolic is the low point, between heartbeats, as the stretched arteries spring back and keep blood moving.
• Pulse pressure is the gap between the two. At 120/80 it’s 40, and at 150/70 it’s 80.

Why it changes with age

Large arteries stiffen as you get older, so they stretch less with each heartbeat. In the Framingham data, systolic pressure rises steadily across adult life. Diastolic pressure rises at first, then falls after about age 50 to 60, and pulse pressure widens. That’s why 150/70 is a common pattern in older adults, and why systolic tells you more about your risk later in life.

Why sodium matters

Your kidneys control how much salt and water stay in your body. More sodium tends to mean more fluid in the blood vessels and higher pressure, more so in some people than others. In a 2023 crossover trial, 213 people ate a week of very low sodium and a week of high sodium. Their systolic pressure ended about 8 mmHg lower after the very low week. Pressure fell in most people, though not in all, and each diet only lasted one week.

Why it moves all day

Pressure rises with activity, stress, caffeine, nicotine, cold, pain, a full bladder or talking, and falls with rest. This is normal, and it’s why a single measurement doesn’t tell you much and guidelines rely on averages. Morning numbers often run highest, and the reasons are on why morning pressure runs high.

From normal variation to a disease

Doctors started treating the number as a target because of three kinds of research.

Long-term cohorts. The Framingham Heart Study gave its first exam in 1948 and enrolled 5,209 people in its original cohort. One of its starting hypotheses was that people with hypertension develop cardiovascular disease faster. The researchers found that risk rose step by step with blood pressure, even at levels once called benign.

Treatment trials. The Veterans Administration trials tested drugs against placebo in men. The 1967 report covered men with diastolic pressure of 115 to 129, and that comparison stopped after about 18 months because the gap in serious events between the groups was so wide. The 1970 report covered 380 men with diastolic pressure of 90 to 114, and treatment came out ahead there too.

Pooled studies. In 2002, researchers pooled 61 studies covering 1 million adults and found no sign of a safe threshold down to at least 115/75. At ages 40 to 69, a difference of 20 systolic or 10 diastolic went with about twice the death rate from stroke and from heart disease. These are patterns across whole populations, so they can’t predict what happens to any one person.

In 2015 the SPRINT trial compared a systolic target under 120 with one under 140 in 9,361 high-risk adults without diabetes. It stopped early after the lower target cut major events by 25 percent and death by 27 percent. Two years later the US guideline moved the line for hypertension from 140/90 to 130/80, which raised the share of US adults with the label from 32 percent to 46 percent. The authors expected only a small rise in the number of people needing medication.

The 2025 AHA and ACC guideline kept those categories and brought in the PREVENT risk equation.

What good looks like

The categories

Whichever of your two numbers sits in the higher band sets your category. Click anywhere on the map to see where any two numbers fall on the US table.

120/80 sits right on a line in the US table, and the diastolic number puts it in stage 1. See what 120/80 means, what 130/80 means and what 140/90 means.

Home against clinic

The American Heart Association lists the matching numbers for each setting.

Targets

The 2025 US guideline gives adults with high blood pressure a goal under 130/80. Ask your doctor which target fits you.

Averages by age

Average blood pressure goes up with age, and so does the risk that comes with it. But the average for people your age doesn’t tell you what’s healthy for you. Guideline categories use the same lines at every adult age. The chart shows averages and categories separately.

When low is a problem

The US National Heart, Lung, and Blood Institute defines low blood pressure as under 90/60. Many people under that line feel fine and don’t need anything done. Lightheadedness, fainting, confusion, blurred vision or a sudden drop from your usual are reasons to call a doctor. See what blood pressure is too low. Cold, clammy skin, unusually fast breathing, a bluish tint or a weak rapid pulse can be signs of shock and need emergency services.

How to measure it

How you sit and which cuff you use can change the number a lot.

Device. Choose an upper-arm cuff from a validated list. The American Heart Association says many home devices on sale haven’t been validated, and that wrist devices are hard to use accurately. How to track lists models.

Fit. Arms measuring 22 to 26 cm take a small adult cuff, 27 to 34 cm an adult, 35 to 44 cm a large and 45 to 52 cm an extra large. In a crossover trial of 195 people, a regular cuff on an arm that needed a large one added 4.8 mmHg to the systolic number, and on an arm that needed an extra large one it added about 19.5 mmHg. Use the cuff size tool.

Setup. Don’t have caffeine, smoke or exercise for 30 minutes before. Empty your bladder. Sit for 5 minutes with your back supported, your feet flat and your legs uncrossed, with your arm on a table at heart height and the cuff on bare skin. Don’t talk.

The week. Take two measurements 1 to 2 minutes apart, morning and evening, for seven days, then average them. The AHA statement recommends at least 7 days and says 3 will do. Follow the seven-day home week and use the home average tool.

What mistakes add. The American Medical Association and Johns Hopkins estimate that talking adds about 10 mmHg, an unsupported arm 10, a full bladder 10, an unsupported back and feet 6, crossed legs 2 to 8, a small cuff 2 to 10, and a cuff over clothing 5 to 50.

White-coat and masked patterns. Some people run high only in a clinic. Others run normal in a clinic and high elsewhere. In a Spanish registry of 59,124 patients, reported in the Lancet in 2023, masked and sustained hypertension both went with a higher risk of death. The researchers only observed these patients, so this shows a link but can’t prove cause. If you only measure in a clinic, you’ll miss the masked pattern. More on both patterns.

24-hour monitors. A device worn for a day and night measures pressure during normal life and sleep. The AHA statement calls it the reference standard for measuring outside a clinic, and calls home measurement the alternative when you can’t get one or can’t tolerate it.

Watches. Apple Watch received FDA clearance in 2025 for a hypertension notification that looks at 30-day periods of sensor data. It doesn’t give a number. In its clearance study of adults without a prior diagnosis, it caught 41.2 percent of people with hypertension and 53.7 percent of those at stage 2, and gave no alert for 92.3 percent of those without hypertension. The FDA summary says that having no notification doesn’t mean you don’t have hypertension, so you still need a cuff to check.

Irregular pulse. The AHA statement says automatic cuffs give a valid systolic number in atrial fibrillation, but a less accurate diastolic one. Ask your doctor how to measure.

What moves it, ranked by effect size

Most of these numbers come from the 2017 US guideline’s table of non-drug changes, and the others come from the studies named in each row. People with hypertension tend to see the bigger drop, and people without it the smaller one. The 2025 US guideline keeps lifestyle as the first line of care for all adults. The full table with time to effect, confidence and exclusions is on the how-to page.

For timing, see how fast you can lower it. For diastolic pressure alone, see how to lower diastolic pressure. Slow-breathing devices showed no clear effect in blinded comparisons. We couldn’t find a supplement that reliably lowers it.

Best practices

1. Measure first. Without a home week to start from, you can’t tell if anything you do is working.
2. Start with the changes that are biggest and fastest for you. Sodium and DASH can show within two weeks. If you drink more than two a day, cutting alcohol is a good early one.
3. Test one change at a time when you can. Repeat the home week the same way after the change has had time to work.
4. Don’t expect the drops to add up. In ENCORE, 144 overweight adults not on medication followed one of three plans for 4 months. DASH alone lowered clinic pressure by 11.2/7.5 mmHg. DASH with exercise three times a week and weight loss of 8.7 kg lowered it by 16.1/9.9. The group that kept their usual diet dropped 3.4/3.8. On a 24-hour monitor the same two active plans gave 5.3/2.9 and 10.2/5.4. Those people had weekly supervised sessions, so you’ll likely get less from the same plan on your own.
5. Keep taking what you’re prescribed. Lifestyle changes go alongside any treatment, and any change to a dose comes from your prescriber.

When life changes the advice

Over 60. Systolic often keeps rising while diastolic falls, so wide gaps are common. Dizziness when you stand up and falls become a bigger concern. Your doctor sets your target.

Pregnancy and after birth. The thresholds are different, and the usual lifestyle advice doesn’t apply. If your pressure was normal before and you’re past 20 weeks, 140/90 or higher on two checks at least 4 hours apart means calling your care team. 160/110 or higher is the severe range. A severe headache or vision change needs care right away. Keep reporting these numbers and symptoms in the weeks after birth.

GLP-1 drugs. In the STEP-1 trial, semaglutide lowered systolic pressure by 5.1 mmHg, alongside a 12.4 percent weight loss. If you already take blood pressure medication, tell your prescriber about dizziness or numbers far below your usual, and never change a dose yourself.

Athletes and lifters. Fit people often run low, and low pressure with no symptoms doesn’t need anything done. Larger arms need a larger cuff or the number comes out high.

Diabetes or kidney disease. The 2025 US guideline uses your overall risk to decide when medication comes up, and your care team sets your targets. The advice on potassium changes too, so ask your care team before you raise it or use a potassium salt substitute.

On blood pressure medication. Your prescriber can use your home averages. Report dizziness on standing and numbers far below your usual. Don’t stop, skip or change anything on your own.

The key studies

What the evidence does not show

• Isometric holds aren’t proven to be the best exercise. The 2023 ranking put them first, but it was based on smaller trials. Keep aerobic exercise as the main part of your routine.
• The drops from each change don’t add up. In ENCORE the combined plan lowered systolic pressure by 16.1, well short of what you’d get by adding up each change on its own.
• The drops seen in trials don’t predict your own result. Many are averages from supervised programs or from studies where food was provided. Your result can land higher or lower.
• Slow-breathing devices haven’t done better than the comparison in trials. When researchers pooled blinded trials against active controls, the systolic difference was too small to be significant. The earlier positive results came mostly from trials the manufacturer was involved in.
• Tighter targets haven’t been tested equally in everyone. SPRINT left out people with diabetes.
• Results for a whole population don’t tell you your own risk. The Ettehad and Lewington researchers pooled results from large groups, so they can’t predict one person’s future.
• Having no watch notification doesn’t rule out hypertension. The FDA’s summary of the watch feature says this too.

How to know it worked

Compare two seven-day averages taken the same way, with the home average tool.

In one analysis of 163 people, a week of home measurements varied less than clinic or 24-hour ones. But even a week only got the average to within about 8 to 10 mmHg of the true figure, with 80 percent confidence. A shift of a few points could just be normal day-to-day change, so look for a change that holds over repeated home weeks. Give each change its time before you retest, about a week or two for sodium and DASH and longer for weight and exercise.

When to see a doctor or call now

Your doctor knows your history and can tell you which of these fits you.

Bring to the appointment: your seven-day sheet with times and the arm you used, your cuff and its model number, every medicine and supplement you take, and how much you drink.