Start with the schedule, then choose the memory size

Memory capacity is printed on almost every box, usually as a number of readings. Current upper arm monitors sold in the United States range from about 14 readings on basic models, to 30 on midrange models, to 60 readings for each of two users, or 90 to 120 readings in total, on the more expensive ones.

Run those numbers against the 28 reading week. A 14 reading memory holds half a week. Once it fills, the oldest readings are overwritten, so by the time you bring the monitor to an appointment the first days of your run are gone. A 30 reading memory holds one full week with a little room to spare. A memory of 60 or more holds two weeks or longer, which matters if you monitor for a week before each visit and want to keep the previous run for comparison.

The 2017 American College of Cardiology and American Heart Association guideline says monitors with memory are preferred, and the 2019 AHA statement explains why. Concern has been raised about the accuracy of recall and written logs, and the statement says a device with built in memory that stores readings automatically addresses that concern. People round numbers, skip an unusually high one, or forget an evening. The memory does none of that.

Two details matter more than the raw number. First, look for a memory that stores the date and time with each reading. A list of 30 readings with no time stamps cannot tell your health care professional which were morning and which were evening, and that separation is part of how home readings are interpreted. Second, if anyone else in your home will use the monitor, look for a two user memory or a guest mode, which lets a visitor take a reading without saving it to your record. In a single user memory, a second person's readings blend into yours and spoil the weekly average. A separate article in this series covers what to ask before sharing a monitor.

Averaging: useful, with one thing to check

Many monitors now advertise averaging. The word covers two different features, and it helps to know which one you are buying.

The first type takes several readings automatically in one sitting, pausing between them, and displays the average. The 2019 AHA statement notes that some devices can be programmed to take multiple readings with a chosen interval between them, such as 1 or 2 minutes. This matches how guidelines want readings taken. The 2025 AHA and ACC guideline says a single reading is inadequate for clinical decisions and calls for an average of two or more readings on two or more occasions. A primary care study of more than 38,000 adults with high blood pressure found that simply repeating the measurement at the same visit lowered the systolic number (the top number) by a median of 8 mm Hg, largely because the first reading tends to run high. A monitor that takes the repeat for you removes the temptation to stop after one.

The second type shows the average of the last several stored readings when you press the memory button. That can be handy, but check what it averages. Some models average the last three readings regardless of when they were taken. Others average a morning set or a full week. The number your health care professional wants is the average of all readings across the whole monitoring period, morning and evening together, and the AMA's guide for health care professionals includes an averaging tool for exactly that. Read the manual or ask before you buy so you know what the average on the screen represents.

What a large display should mean for you

Mayo Clinic's advice is plain: the display should be clear and easy to read. Manufacturers use phrases like "large LCD" and "extra large display," but there is no standard behind those words. One widely sold talking model lists its screen at 1.7 by 1.9 inches, roughly the size of a matchbook, while other brands call a screen "large" without giving dimensions. The only reliable test is to look at a display in person or ask a pharmacy to open a box.

When you look, judge four things: whether you can read the systolic and diastolic numbers (the top and bottom numbers) from arm's length without leaning in, whether the pulse is clearly separated from the blood pressure numbers so you do not mistake a heart rate of 72 for a diastolic reading, whether there is a backlight or strong contrast for a dim bedroom in the early morning, and whether the screen shows the date and time alongside each stored reading when you scroll back through memory. The first point matters more than it seems. Leaning forward or twisting to see the screen is movement, and movement is one of the common reasons a reading is thrown off or flagged with an error.

A few symbols are worth understanding before you buy. An irregular heartbeat symbol means the monitor detected an uneven pulse during the reading. It is not a diagnosis, but a symbol that appears repeatedly is worth mentioning to your health care professional. A movement symbol means you shifted during inflation, and a cuff fit symbol on some models means the cuff was wrapped too loosely. Both tell you to repeat the reading rather than trust it.

A color coded category indicator deserves a note of its own. Some models light an indicator when a reading is 130 or higher on top or 80 or higher on the bottom. Those cutoffs match the range the 2025 guideline calls stage 1 hypertension. The light is a device feature, not a diagnosis. Whether you have high blood pressure depends on the average of repeated readings over days, interpreted by a health care professional who knows your history. If you are someone who worries at a flashing color, a model without an indicator, or one that lets you turn it off, may serve you better. The European Society of Hypertension's 2023 guideline notes that home monitoring can lead to anxiety and overly frequent measuring in some people.

Simple controls, and monitors that talk

The AMA's position is that a home monitor should have an upper arm cuff and inflate automatically, and the 2025 guideline treats automated devices as reasonable in place of the stethoscope method. Manual squeeze bulb monitors take training to use well and have no memory. Beyond that, the fewer buttons you need for a normal reading, the better. Most good monitors need one press to start and inflate, one to stop, and one to scroll memory.

Watch for controls that are hard to use with the arm you are not measuring. Guidelines call for the cuffed arm to rest supported on a table, so you operate the monitor with your other hand alone. Buttons that are small, flush, or crowded together are harder for stiff fingers. If arthritis, tremor, or weakness affects your hands, press the buttons in the store.

The cuff is a control too. A cuff with a rigid ring can be threaded and pulled snug with one hand, which is easier than lining up a soft wrap cuff by yourself, and preformed cuffs hold a curve that some people find easier to slide on. Whatever the style, the printed range on the cuff must include your arm. A separate article in this series explains how to measure your arm before you buy.

For people with low vision, monitors that speak exist. One model sold in the United States gives spoken instructions and reads the results aloud in English, Spanish, or French, and the voice can be turned off when it is not wanted. Talking monitors also help people who find it awkward to hold still while trying to read a screen, and they let a family member in the next room hear the result. Independent guidance on talking monitors from health organizations is thin, and the feature is not on any guideline list. Treat it as a practical aid, and hold the model to the same standards of validation and cuff fit as any other.

Batteries, adapters, and warranties

Nearly all upper arm monitors run on four AA batteries. One manufacturer states that its talking model gets about 700 readings from a set of alkaline batteries under test conditions. At the AMA schedule of four readings a day, that is roughly six months of daily use, and far longer for someone who monitors for a week before each visit. Low battery symbols appear well before the batteries die; one manual says its symbol starts flashing when about three quarters of the charge is used. Manuals agree on the basics: replace all four at once, do not mix old and new, and take the batteries out if the monitor will sit unused for months. A separate article in this series covers cleaning and storage.

Check whether batteries are included and whether a wall adapter is included, sold separately, or not offered. If the monitor will sit in one place, an adapter saves money over the years. At least one manufacturer states that its monitors run on wall power without any batteries installed, which suits a monitor that lives on a nightstand.

Warranties on the monitors reviewed for this article ranged from 2 to 5 years on the unit and 1 to 2 years on the cuff. The cuff wears out first. Manufacturers state service lives such as 5 years or 10,000 measurements for the unit and 2 years or 5,000 measurements for the cuff, figured at about six readings a day. A replacement cuff usually costs less than a new monitor, so before buying check that the brand sells cuffs separately in your size. Warranty terms are often found only in the instruction manual, not on the box, so read the manual online before purchase if you can. A separate article in this series covers how to find the instructions for your monitor.

Both heart.org and Mayo Clinic suggest bringing the monitor to your health care professional's office about once a year, or as the manufacturer directs, to compare its readings with the office equipment. Some manuals recommend an accuracy check every 2 years or after the monitor has been dropped.

Bluetooth and apps: convenient, not required

Connected monitors send readings to a phone app over Bluetooth, where storage is effectively unlimited and trends appear as charts. Many apps can export a report or share readings through a clinic's secure portal. The 2025 guideline recommends home monitoring for people taking blood pressure medication along with supports such as education and telehealth, and electronic sharing makes telehealth easier.

None of this is required. The AMA notes that readings can reach your care team as a written log, as device memory you bring to the office, or through connected apps and secure portals. Before paying for Bluetooth, consider three practical things: whether you have a phone that will run the app and a habit of keeping it charged and updated, whether you or a family member who helps you will actually open the app and send a report, and whether you are comfortable with the app's privacy policy. If any of those is a no, a monitor with solid memory and time stamps does the job, and the office visit with the monitor in hand does the rest.

Smartwatches and other cuffless wearables that estimate blood pressure are a different category. The 2025 guideline advises against relying on them for diagnosis or treatment decisions. A Bluetooth cuff monitor and a smartwatch estimate are not the same product, even when the same app displays both.

Price, coverage, and a shopping list

On one manufacturer's current US price list, validated upper arm monitors run from about $40 for a basic model to about $55 for a small cuff model, $60 to $80 for two user and talking models, and about $155 for a model built around an extra large cuff. A 2020 study of nearly 1,000 monitors sold online in Australia found that validated devices cost more on average than nonvalidated ones, so the cheapest listing is often not a validated one. Validation, covered in a separate article in this series, comes before any feature on this page.

Medicare's list of covered durable medical equipment on medicare.gov does not include home blood pressure monitors. Medicaid coverage varies by state. Mayo Clinic notes that some insurance plans cover a monitor, so a call to your plan or a question at the pharmacy counter is worthwhile before you assume you are paying out of pocket.

When you are ready to buy, take this list with you.

1. 1 An upper arm cuff whose printed range includes your measured arm circumference, with a replacement cuff available in that size.
2. 2 Validation confirmed for the exact model number, not just the series name.
3. 3 Automatic inflation and a one button start.
4. 4 Memory of at least 30 readings, each stored with date and time, plus a two user memory or guest mode if the monitor will be shared.
5. 5 An averaging function you understand: either automatic repeated readings in one sitting, or a stored average whose time span you have checked.
6. 6 Digits you can read at arm's length without leaning, a backlight if you read in dim light, and a voice option if your vision is limited.
7. 7 Buttons you can press comfortably with one hand.
8. 8 Four AA batteries included or a wall adapter you have priced, and a note on whether the unit runs on wall power alone.
9. 9 Warranty terms for both the monitor and the cuff, ideally 2 years or more on the unit.
10. 10 Bluetooth or an app only if you will use it, and never a cuffless wearable in place of a cuff.

Suppose your eyesight is fair, you live alone, and your health care professional has asked for a 7 day run of readings before each visit. Your list narrows fast: a validated single user monitor with 30 or more time stamped readings, large digits, and an adapter, at roughly $40 to $60. Suppose instead that you and a partner both monitor and one of you has a much larger arm. Now you need a two user memory and possibly two cuffs. The features that matter are the ones your own routine will use every morning and evening.