How to Choose an SD or microSD Card: Device Compatibility, Capacity and Speed

Read the package labels and you're 90% there — first, sort out the "layers"

Look at any SD card shelf or online listing and you'll see labels like "microSDXC," "UHS-I," "U3," "V30," "A2" and "150MB/s" crammed onto the card. This is what makes it confusing, and the main reason people freeze up wondering which one to buy. But these labels are just different kinds of information sharing one card, and once you read them by type, it all falls into place. Let's start with the two base layers: "size" and "capacity standard."

Size: SD and microSD are different, but you can bridge them

The two main sizes are standard "SD" and the fingernail-sized "microSD" (the in-between miniSD has basically disappeared). Roughly speaking, DSLRs, mirrorless cameras and full-size camcorders use standard SD slots, while most phones, Nintendo Switch, dashcams, GoPros and drones use microSD slots. The key thing to remember: a microSD card works in an SD-slot device if you put it in the included or a store-bought adapter. The reverse — putting a standard SD card into a microSD slot — is physically impossible. So if in doubt, buying one microSD card and using it with an adapter gives you the most flexibility.

Capacity standards: SDHC, SDXC and SDUC are "different ceilings"

Even at the same size, the name changes depending on the maximum capacity. These are easy to mix up, so here's a table.

The point: unless your device says "SDXC compatible," a card of 64GB or more may not be recognized or may not work properly. Many cases of "I bought 256GB to be safe, but my old car navigation system (or some other device) couldn't read it" come down to overlooking this.

Breaking down the "four-way" speed labels — what to check so recording doesn't stop

Speed is the most confusing part of SD cards. One card can carry up to four speed ratings at once — Speed Class, UHS Speed Class, Video Speed Class and App Performance Class. On top of that, the big "XXMB/s" printed on the package is usually the maximum read speed under ideal conditions, which is not the same as the "minimum write speed" that actually matters when recording. Mix these up and you get "it's supposed to be fast, but 4K recording keeps stopping."

Shooting video vs. running apps: you look at different marks

Which speed mark matters depends clearly on what you're doing. For video and burst shooting, you need the ability to keep writing data without interruption, so look at the V class (or U3). For running apps or games on a phone or Switch, what you feel is how fast it handles lots of small reads and writes rather than writing big files in one go, so A2 (or A1) matters more. Even at the same capacity and price, a card optimized for video and one optimized for apps have different strengths.

"Bigger is always better" doesn't apply — device limits and the "all your eggs in one basket" problem

It's tempting to buy the biggest card your budget allows, but with SD cards, large capacities come with two specific pitfalls.

Pitfall 1: Cards over the device's "maximum supported capacity" won't be recognized

Every device has a maximum supported capacity. For example, put a 512GB card into a device whose manual says "up to 256GB" and you may find it isn't recognized, shows the wrong capacity, or only partly works. This is decided by both the standard (SDXC support or not) and the maximum capacity, so the golden rule is to check "compatible cards" and "maximum capacity" in the device's spec sheet or manual before buying. Older phones, car navigation systems and entry-level digital cameras are the most likely to trip you up here.

Pitfall 2: Put everything on one card, and you lose everything when it fails

SD cards are consumables and can suddenly become unreadable without warning. Keeping a thousand travel photos on a single 512GB card, only to have it stop working on the last day, is not a rare story. The more you consolidate onto one card, the bigger the loss when it fails. Professional photographers deliberately use several mid-size cards instead of one huge one to spread the risk. When it comes to protecting your data, "bigger is better" isn't necessarily right.

The "right pattern" for each device — the same card can be a good or bad fit

The best SD card changes a lot depending on what you're putting it in. Here are the marks to check and the cautions for common uses.

Extra storage for a phone

As a place for photos, music and app data, an A2-rated microSDXC (128–256GB) is the easy standard choice. If you sometimes run apps from the card, A2, which guarantees random performance, makes a noticeable difference. Two things to check: whether your phone has a microSD slot at all (many recent models don't), and its maximum supported capacity. Always check before inserting.

Nintendo Switch

As your downloaded games pile up, internal storage fills fast. A large microSDXC (128–512GB) is the basic choice, and since it affects game loading, some speed (UHS-I, A class) makes it more comfortable. Choosing a fast, large card that meets Nintendo's recommended conditions is the safe bet. Note that if you insert an exFAT card, the Switch may ask for an update, so follow the console's own formatting process.

Digital cameras, DSLRs and mirrorless cameras

For mainly stills, C10/U1 is enough, but the safe lines are V30 or higher for 4K video, and V60/V90 plus UHS-II for fast burst shooting, 8K and professional work. What matters here is whether the camera supports UHS-II — put a UHS-II card in a camera without support and it only runs at UHS-I speed. For important shoots, as mentioned in the previous section, recording across multiple cards reduces the risk of losing every shot to one failed card. If your camera has dual slots, use simultaneous (backup) recording too.

Dashcams and security cameras

This use is in a class of its own: because the card keeps overwriting the same area endlessly, ordinary cards wear out quickly and start throwing errors or failing to record. Always choose a dashcam/surveillance model labeled "high endurance", and regularly check that it's recording properly and replace it. The only way to avoid "it didn't capture the crucial moment" is not to trust it blindly and to check it. Match capacity to recording time and overwrite settings — too large can keep overwriting from working properly, so follow the device's recommendation.

GoPro, action cameras and drones

Since you're shooting high-bitrate 4K/5K video outdoors in tough conditions, the key is V30 or higher (V60 recommended for some models) plus stable writing. Manufacturers often publish lists of "tested cards," and to avoid compatibility issues, choosing a model number from that recommended list is the reliable approach. A design that handles temperature swings, shock and water gives extra peace of mind.

Why "up to 160MB/s" and "V30" appear on the same box — read each number by where it comes from

SD card product pages are confusing because numbers with different meanings sit on the same line. Catalog maximums, minimums guaranteed by the standard, and theoretical bus (connection) limits — three numbers from different sources are lumped together as "speed." Just separating them makes comparing much easier.

Max read / max write are "records set on a good day"

The big "up to XXMB/s" on the front of the package is usually the theoretical peak read speed. Think of it as measured with large continuous files, a compatible reader and bus, and caching in effect. What matters for shooting is the write speed, and when writing lots of small files it doesn't reach the advertised number. In practice, if a card doesn't list a max write speed, assume writing isn't its selling point.

Marks are a "floor promise"; numbers are a "ceiling expectation"

Marks such as C10 (Class 10), U1/U3 and V6–V90 are all declarations of a floor: the card will sustain at least this write speed. U3 and V30 both mean a floor of about 30MB/s, so having both on a card doesn't double the speed. With V60 and V90, the floor itself goes up, so "recording stopped because writing couldn't keep up" is less likely with high-bitrate video. For devices that record continuously, look at the floor, not the ceiling — that's the most important trick for reading speed labels.

A1/A2 are for apps, not video

The A1 and A2 marks show the minimum random read/write performance (IOPS) when apps are stored on the card in a phone or single-board computer. This is a separate axis from sequential write speed and doesn't guarantee stable video recording. Conversely, a V60 card sold for cameras may not even claim A1. If your use is running apps, look at A2; if it's recording, look at the V class — switch which mark you check from the start and you won't get lost.

UHS-I / UHS-II / SD Express are "the number of lanes on the road"

One row of contacts on the back of the card means UHS-I; two rows means UHS-II. A UHS-II card won't break in a UHS-I-only device, but the second row isn't used and speed drops to UHS-I level. More recently there's also SD Express, a standard that runs PCIe/NVMe, with a further change in contact layout. Most cases of "I bought a fast card but it isn't fast" are caused not by the card but by the number of lanes on the device or reader side.

Also, a "64GB" card showing up as about 59GB on your device is due to the difference between counting 1GB as 1 billion bytes or as 2 to the 30th power — it's neither a defect nor a fake. Remember that a drop of just under 10% is normal, and you'll avoid unnecessary returns.

UHS-II or UHS-I, or offload to an SSD — three common forks in the road

What people actually struggle with when choosing an SD card isn't so much "what the standards mean" but whether it's worth paying for the next grade up. Here are three commonly compared pairs, broken down by conditions.

UHS-II is worth it for "after a burst" and "transfers"

With its second row of contacts, UHS-II sits in a higher price range. Where it helps is clear: one is how quickly the camera recovers after the buffer fills during burst shooting, and the other is transfer time with a compatible reader. On the other hand, if you're just recording video at a constant bitrate, a UHS-I V30 card is enough on paper. If the camera itself doesn't support UHS-II, the second row is physically unused, so the money only buys you faster transfers. Some cameras have two slots, with only one supporting UHS-II, so check the per-slot support in the camera's spec sheet before buying.

Whether V30 is enough or you need V60+ — work it out from the bitrate

You can decide by calculation, not gut feeling. If the camera's recording bitrate is in Mbps, divide by 8 to get MB/s — for example, 200Mbps is about 25MB/s. Then add a margin: an easy approach is to pick a class whose floor is roughly 1.5 times what you actually need. Standard 4K recording often fits within V30, while ALL-Intra, 6K, 8K and RAW video move into V60 and V90 territory. Some cameras state "V60 or higher required" in their specs; if yours does, it's safest to just follow it.

One large card or several mid-size ones

Using one card means less swapping and fewer dropped cards, but the moment that card becomes unreadable, everything is gone. The less you can reshoot something, the more you should protect it with more cards rather than more capacity.

Not relying on SD at all

These days, cameras that record directly to an external SSD, and offloading phone photos via cloud sync, have become common. For long high-bitrate recording, SSDs handle capacity more easily and spread heat better. On the other hand, the extra power and cables reduce mobility, and there's the risk of loose connections outdoors. For environments with vibration or frequent insertion/removal, like handheld shooting or dashcams, cards are the simpler choice. If you rent a CFexpress-compatible body as pro gear, check in advance whether cards can be rented too (and whether media costs extra), so you don't end up on the day with "nothing to put in it."

What "high endurance" means, and how to deal with card lifespan

The term "high endurance" (高耐久) on card packaging is often chosen on vibes, but it means something specific: being resistant to repeated writes, as in continuous recording, and, depending on the product, being physically tough — waterproof, temperature-proof, shockproof, X-ray-proof and magnet-proof. Flash memory wears a little with every write, so for dashcam and security uses that write to the same area endlessly, whether a card has this durable design directly affects its lifespan.

Signs of wear and when to replace

SD cards can fail "out of the blue," but they often show warning signs first. If you start seeing more of the following, it's time to replace the card.

• Writing or reading has become noticeably slower: especially waiting on video transfers or saves
• Errors or "Please format the card" messages appear: the data management area may be starting to degrade
• Files are more often partly corrupted or won't open: the card may be near the end of its life, or there's a compatibility/contact problem
• A dashcam or security camera has gaps in recording: replace early, especially for always-writing uses

With prices now reasonable, the safest approach for cards holding important recordings is to replace them with new ones regularly, before symptoms appear. For continuous recording, keeping a spare card lets you swap it in right away when needed.

Avoiding fakes and fake-capacity cards — how to spot "super-cheap huge cards"

Because SD cards are small and hard to tell apart, shoddy fakes with faked capacity slip into online stores easily. A fake-capacity card shows "512GB" on your device but actually holds far less; write beyond its real capacity and data gets corrupted, or the first files you saved get overwritten and lost. You could discover too late that all your precious photos are gone.

Checklist for spotting and preventing fakes

• Be suspicious of large cards priced far below market: a price like "1TB for a few hundred yen" is a clear sign to suspect fake capacity
• Choose genuine products from trusted brands: SanDisk, Kioxia (formerly Toshiba), Samsung, Lexar, etc., from sellers with a manufacturer warranty
• Check the seller and warranty: official manufacturer or brand store, long-term warranty or not. Avoid sellers with odd-looking names
• Check the real capacity when it arrives: use a capacity-testing tool to confirm it can be written all the way to the stated capacity
• Rough packaging or printing is also a clue: blurry logos, mismatched model numbers and overly plain packaging are red flags

*Brands and series are only examples. Even within the same brand, parallel imports or unofficial products can creep in depending on where you buy, so the biggest tip for avoiding fakes is to factor in where you buy from when you choose.

Buying for less — timing and online stores, SD card edition

SD cards are a category where the price per GB keeps falling as technology improves, so rather than waiting for "someday it'll be cheaper," the basic rule is to buy when you need one. Still, for bulk purchases or upgrading to a larger card, timing and store choice can lower your real cost. Current prices change, so here we'll focus on how to think about when and where to look.

What to look for at each store, for SD cards

• Amazon: lots of reviews, making it easy to find "comments from people using it with the same device as you (a specific GoPro or dashcam model, Switch, etc.)". Search the reviews for your device name to find impressions on compatibility and recording stability — useful even for cards not on a recommended list. Also check whether it's "Ships from Amazon" or an official store
• Rakuten: easy to stack points with Shopping Marathon and SPU (Super Point Up program), but price, shipping and warranty differ by shop even for the same product. Compare the effective price including points, shipping and warranty length so you're not fooled by the apparent lowest price
• Yahoo! Shopping: some days have PayPay-related rewards, and lining up effective prices including rewards can work in your favor. If an official electronics-retailer store is listed, you also get the reassurance of a genuine product

The common rule at any store: first decide whether the capacity and speed class are enough for your use, then compare effective prices between cards with the same specs. "It only looked cheaper because it was one speed class lower" is a common mistake, so compare V30 with V30 and A2 with A2 to avoid it.

FAQ