Why is my blood pressure different every time I check it?

Your blood pressure is different every time because it really does change from moment to moment, with every heartbeat, breath, and change in posture. On top of that, small slips in how you measure can push a single reading up by anywhere from 4 to about 20 points, which is often bigger than the real change you're trying to see.

So when you get 128 and then 121 a minute later, neither number is wrong. Each one is a snapshot. Here's how to tell normal scatter from fixable error, and when the swings themselves are worth a conversation with your doctor.

Three reasons your numbers move

When readings jump around, it's usually one of three things, and they call for different responses:

1. Normal biological variation. Your body adjusts blood pressure constantly. You can't remove this, and you shouldn't try.
2. Measurement error. Posture, cuff size, arm position, talking, and timing. You can fix almost all of this.
3. Variability that may carry information about risk. Large swings over months and years, which researchers are still learning how to use.

Normal variation, and how big it is

The American Heart Association's 2019 statement on blood pressure measurement says perfect reproducibility isn't possible, because pressure varies from one heartbeat to the next. Its home monitoring checklist even tells clinicians to warn patients that individual readings "may vary greatly (high and low)" across a monitoring period.

Some numbers on how much:

• The first reading often runs high. The same AHA statement cites a US study in which 35% of adults whose first office reading was 140 to 159 over 90 to 99 came in under 140/90 once three readings were averaged. There's more on this in why your first blood pressure reading is higher.
• Day to day, at home. In a Swedish study of 454 people on blood pressure treatment who measured every evening for 8 weeks, systolic pressure varied from day to day by an average of 7% of each person's mean. With a home average around 135, that works out to roughly 9 to 10 points of typical scatter, even with no change in health or treatment.
• Across time of day. Mornings and evenings can differ too, and the size of that gap varies between populations. See morning vs. evening blood pressure.

Measurement error you can fix

These slips push readings in a predictable direction, mostly up. The numbers below are for systolic pressure, the top number.

The AHA statement also lists talking or texting, a full bladder, a cuff over clothing, and having caffeine, exercising, or smoking within 30 minutes of the reading. It doesn't put a single number on each of those, so I won't either.

Stack two or three of these and you can easily add 15 or 20 points. Someone on the couch with their arm on their lap, legs crossed, chatting, in a cuff a size too small, is mostly measuring their setup. The fix is the same every time: 5 quiet minutes seated, back supported, feet flat, cuff on a bare upper arm, arm resting on a table at heart level. How to measure blood pressure at home walks through it step by step.

Variability that may matter clinically

Researchers have looked at whether bigger swings, especially between clinic visits over months or years, signal higher risk. A 2016 BMJ meta-analysis pooled 19 observational cohorts and 17 clinical trial cohorts. Higher long-term variability in systolic pressure was linked to higher risk independent of average blood pressure. For each standard-deviation step up in variability:

The authors also flagged limits. About half the studies they reviewed had methodological problems (the link held in the studies that avoided them), results varied between studies, and they wrote that variability "is not easily assessed clinically" and that it's unclear which way of measuring it is best. This isn't something to calculate at home. If your readings swing widely week to week even with good technique, that's worth showing your doctor, and a log of every reading is the best way to show it.

Why averages are what guidelines use

Because any single reading carries this much noise, the AHA's home monitoring protocol is built around averaging. Take two readings at least a minute apart every morning before blood pressure medication and every evening before bed, ideally for 7 days (28 or more readings; 3 days is the minimum). Then average all of them. Some guidelines drop the first day.

A 2010 study in the British Journal of General Practice shows why. Within-person variation was 8.6% for office readings, 5.5% for a 24-hour monitor, and 4.2% for a week of home readings. The authors estimated it would take 10 to 13 office readings to pin blood pressure down precisely enough to fine-tune medication, and they found a week of home monitoring the most accurate of the three methods, though still imperfect.

How much difference is normal?

Here's what those studies suggest, translated for someone whose systolic pressure averages about 135. Your numbers will scale up or down with your own average.

These are averages across groups of people, so treat them as a rough sense of scale. Some people bounce around more than others.

A few practical rules follow from them:

• Two readings a minute apart that differ by 5 or 10 points sit well within that scatter. Average them.
• One reading that's 15 points off your usual, with nothing else going on, is most likely noise or a setup problem. Check your posture and cuff, rest, and measure again.
• A weekly average that moves by more than about 5 or 6 points, and stays there for a couple of weeks, is more likely a real change.
• If back-to-back readings regularly differ by 20 points or more despite good technique, bring your monitor to your next appointment so your clinic can check it against theirs.

When to call your doctor or 911

Call your doctor if your weekly averages climb into a higher category and stay there, if your readings swing widely even when you measure carefully, or if low readings come with dizziness or lightheadedness when you stand up.

Call 911 for any high reading that comes with chest pain, shortness of breath, vision changes, slurred speech, weakness or numbness on one side, or a sudden severe headache. Don't wait to recheck in that situation.

Common questions

Why is my second reading lower than my first?

The first reading often runs higher, which is part of why guidelines ask for two readings a minute apart and use the average. If the second is a few points lower, that's expected.

Is a 10-point difference between readings normal?

Often, yes. In a Swedish study of people measuring at home daily, typical day-to-day scatter was around 9 to 10 points systolic for someone averaging 135. Check your technique, then look at the weekly average.

Which reading should I believe?

None of them alone. The AHA's home protocol averages all readings over about a week, and that average is the number your doctor will want.

Should I keep measuring until I get a normal number?

No. Picking the lowest reading skews your log downward, the same way one rushed reading skews it up. Take your two readings, average them, and move on with your day.

How BP Log helps

BP Log's guided flow times two readings a minute apart and averages them for you, following the 2025 AHA/ACC guideline. You can add extra readings and exclude one from the average if you know it was off, and the dashboard keeps separate morning and evening averages so a week of scatter turns into one number you can actually use.